At the International Conference on Equine Exercise Physiology (ICEEP7), about 70 people attended the workshop on workload and conditioning guided by the authors. Most of the audience were involved in Thoroughbred or Standardbred racing, and only a limited number of people were mainly involved in FEI equestrian disciplines (sport horses). The workshop and this review article address the measurement of workload and conditioning of the Thoroughbred racehorse. It was proposed that workload could be quantified using a few selected parameters commonly recorded in the racing industries, such as velocity and distance, to generate a cumulative workload index. The review of conditioning focuses on the Thoroughbred racehorse and examines what can be modified with training, how training programmes should be designed based upon scientific methods and how training programmes should be routinely designed in current practice. It would appear that, in general, the methods used in practice for training Thoroughbred racehorses are quite similar to those used in a set of recent scientific studies, particularly in young (2–3-years-old) Thoroughbreds. Nevertheless, both the length of the training programme and the total amount of exercise are usually shorter/lower than ideal in order to maximize physiological adaptations within the animal's body. In planning the training programme, it is very important to recognize that different adaptations occur at different rates, and this will affect the relative amount of training that should be applied to achieve specific adaptations.

An optimally functional musculoskeletal system is crucial for athletic performance and even minor perturbations can limit athletic ability. The introduction of the muscle biopsy technique in the 1970s created a window of opportunity to examine the form and function of equine skeletal muscle. Muscle histochemical and biochemical analyses have allowed characterization of the properties of equine muscle fibres and their influence on, and adaptation to, physical exertion. Analyses of exercise responses during standardized treadmill exercise and field studies have illustrated the role of cellular energetics in determining athletic suitability for specific disciplines, mechanisms of fatigue, adaptations to training and the affect of diet on metabolic responses. This article provides a review of the tools available to study muscle energetics in the horse, discusses the muscular metabolic pathways and summarizes the energetics of exercise.

Feeding regimens for horses are usually based on perceived workload (PW) together with body condition. This can lead to inappropriate energy intake and, in particular, excess weight. Therefore, a more detailed understanding of the factors influencing individual energy needs under practical field conditions would be valuable. The purpose of this study was to evaluate to what extent the variations in actual versus PW may influence the estimated energy intake required to maintain body weight (BW) in individual animals. Eight mature, experienced riding school horses, chosen at random and maintaining a constant BW, were observed in standard flat work lessons with four different instructors at an equine training college. Heart rate (HR) and the structure of the lessons were recorded. Relative workload (RW) was determined for each horse by dividing its average HR per lesson by its estimated maximum HR (HRmax). PW was estimated by each instructor for each horse using a scale of 1–5. One-way ANOVA and t-test were used to determine the differences and Spearman's and Pearson's correlations were used to determine the correlations. The mean estimated RW was 39.1% of the estimated HRmax ( ± 5.7%) and was not significantly different between instructors (P>0.05). Rider weight as a percentage of BW showed a weak but statistically significant positive correlation with mean HR (r2 = 0.14, P < 0.05). Estimated energy intake was negatively correlated with BCS (r2 = 0.65, P < 0.001) and differed significantly (P < 0.05) between light (BW = 455–532 kg) horses (mean 0.18 ± 0.04 MJ kg− 1 BW) and heavy (BW = 622–660 kg) horses (mean 0.15 ± 0.03 MJ kg− 1 BW). No difference was found in estimated energy intake between different workloads (flat work only or flat work and jumping) (P>0.05). When BW was compared with estimated energy intake, no relation was found either (P>0.05). These results suggest that the amount of energy needed to maintain BW in individual horses cannot necessarily be estimated simply on the basis of the intensity and the duration of structured exercise. Other factors including age, individual digestive and metabolic efficiency, body condition and possibly the energy utilized through non-structured activity (e.g. box walking) may need to be taken into account.

Silicon appears to be an essential element for proper endochondral ossification and glycosaminoglycan synthesis. Osteochondrosis (OC) is characterized by disturbed endochondral ossification, and glycosaminoglycan loss has been reported in OC lesions. Therefore, silicon may be effective in reducing OC lesions. The objective of the study was to evaluate the effects of supplemental silicon on OC lesion size. Initial radiographs were taken on 2-year-old Standardbreds (n = 44) on private facilities to identify OC defects in the distal third metacarpus/tarsus or osteochondral fragments at the dorsal aspect of the joint and defects/fragments at the distal tibia and on the trochlear ridges of the talus. Only clinically sound horses were included. Follow-up radiographs were taken at 120 days. Radiographs were digitized and the length and height of the OC defects/fragments were measured. Horses meeting study-inclusion requirements (n = 8) were pair-matched by facility and affected joint(s) and assigned to a group: control (receiving 200 g whole grain flour) and silicon supplemented (receiving 200 g bioavailable silicon source). Treatments were top-dressed on the feed for 120 days. Due to the onset of lameness, four horses did not complete the 120-day study duration. There was no effect (P>0.22) of treatment, time or treatment × time on lesion length, height or area. Silicon supplementation did not alter the radiographic appearance of OC lesions. Future research should focus on the effects of a silicon supplement on the prevention of OC lesions.

Despite the high athletic demands of Greyhound racing and recommendations for the inclusion of pre-race warm-up by some veterinarians, the extent of warm-up practices by Greyhound owners and trainers remains unexplored. We postulated that little pre-race warm-up occurred and any warm-up activities were performed at random. A cross-sectional study design was used to examine pre-race warm-up practices in a sample of 80 Greyhounds. Warm-up components were recorded and scores were assigned in order to quantify the total amount of warm-up undertaken. Associations between total warm-up and signalment and race characteristics were analysed statistically. Correlations between total warm-up performed and all variables including race distance, race order, grade of Greyhound and trainer were poor (all r < 0.2). No statistically significant association between total warm-up and any variable examined could be demonstrated at the P = 0.05 level, although a weak trend towards graded Greyhounds performing more warm-up was noted (P = 0.09). We conclude that little pre-race warm-up is performed in Greyhounds and that when practised, the amount and methods of warm-up are inconsistently applied.

Kinematic studies of western pleasure horses are limited and were performed before current changes in the stock breed association judging standards on the western pleasure gaits. The objective was to measure the kinematics of the jog and lope of the stock breed western pleasure horse. Reflective markers attached along palpation points of the joint centres of the left forelimb and hind limb of four stock breed western pleasure horses were tracked for five strides for each gait for each horse. Both the jog and lope were determined to be four-beat stepping gaits. During the jogging stance, the elbow (159.7 ± 6.6°), carpal (179.9 ± 1.1°), fore (227.6 ± 2.7°) and hind fetlocks (227.4 ± 6.9°), stifle (159.5 ± 6.5°) and tarsal (166.5 ± 6.5°) joints demonstrated peak extension. The same joints demonstrated during swing peak flexion with the hind fetlock joint having double peaks of flexion (195.7 ± 3.2°, 182.3 ± 2.1°). During loping stance, the elbow (153.4 ± 4.2°), carpal (179.7+0.4°), and fore (228.3 ± 9.7°, 229.8 ± 10.0°) and hind fetlock (232.1 ± 2.6°) joints of the leading limbs demonstrated peak extension with tarsal peak extension (157.0 ± 9.6°) occurring at lift-off. Peak flexion occurred during swing for the elbow (105.1 ± 3.1°), carpus (119.8 ± 6.1°), hip (83.5 ± 5.4°), stifle (129.8 ± 9.6°) and tarsus (127.5 ± 6.1°). Kinematic measurements will assist in objectively defining the stock breed western pleasure gaits.

The interactions between nutrition, energy status and acid–base balance during exercise are poorly understood. Exercise, under conditions of prior glycogen depletion (GD) and low-carbohydrate diet, results in a decreased rate of skeletal muscle glycogenolysis, greatly decreased muscle pyruvate and lactate contents with decreased plasma [lactate] (Putman et al., Am J Physiol, 265: E752, 1993). Therefore, it is hypothesized that exercise in GD, compared with normal (NG) or high-carbohydrate conditions, will result in a reduced magnitude of acidosis due to reduced production and accumulation of lactate. In two trials (GD, then NG) separated by 1–2 weeks, four men cycled at 75% of peak VO2 until the time of exhaustion in GD (57 ± 7 min). At 2 min of exercise, femoral vein (fv) plasma [H+] was increased by 21 ± 4 neq l− 1 (NG) and 14 ± 3 neq l− 1 (GD); increases in arterial [H+] were only c. 45% of those in fv plasma. The increase in fv PCO2 (NG, 25 ± 2 mm Hg and GD, 15 ± 2 mm Hg) was the primary variable responsible for the increased [H+]. During NG, the increase in fv [lactate− ] exceeded the decrease in strong ion difference [SID], with electrolyte charge balance mainly due to increased [Na+]. In the GD trial, arterial [SID] decreased and was the primary contributor to the increased [H+], as passage of blood through the lungs eliminated the CO2 contribution prevalent in fv plasma. Throughout GD, plasma [lactate− ] increased less than in NG and the decrease in [SID] in GD was also significantly less than in NG. In summary, in GD conditions, an attenuated production/release of lactate−  and CO2 from muscle resulted in reduced magnitude and duration of acidosis compared with NG conditions. In fv plasma, increased PCO2 was the primary variable responsible for the rapid and sustained elevation in [H+], whereas in arterial plasma decreased [SID], due to increased [lactate− ], was primarily responsible for increased [H+].

Regular monitoring of the heart rate (HR):speed relationship may help evaluate response to training and aid in the early detection of problems. This relationship is normally determined using a treadmill or via a ridden test conducted outside on a track. Simple practical alternative methods to obtain this relationship without access to a treadmill or a track could be of value in the field. To evaluate whether the HR:speed relationship could be determined via an indoor ridden test or a lunge test, HR was monitored on two occasions at least 3 h apart, in 12 adult horses (mixed breed) in a familiar environment during a 5 or 7 m radius circle lunge (unridden) test (5LT or 7LT) and an incremental (ridden) test (RT) on the same day. The RT comprised two ridden laps of the perimeter of a 60 × 40 m indoor school at walk, three laps at trot, three at medium canter and four at fast canter (all on the right rein). The speed of each lap was recorded. The LT comprised lunging for 2 min on each rein at walk, trot and canter. Speed was determined from the number of laps completed and measurement of the distance travelled. HR and speed were highly correlated in both lunge and ridden tests (both r = 0.99 ± 0.01). V140 on the ridden test (5.2 ± 0.6 m s− 1) was significantly greater than on the pooled lunge test data (4.4 ± 0.6; P < 0.0001). There was a negative correlation between recovery HR at 2 min following either the LT or RT and V140 (P < 0.05). The slope of the HR versus speed relationship and V140 were not different between RT and 7LT, but were significantly different from those of the 5LT (P < 0.05). V140 was always lower on the lunge tests compared with the ridden test. This suggests that, in this study, lunging without a rider increased the metabolic demand above that for being ridden at a similar speed. V140 determined by the 7LT gave the closest approximation to the V140 determined by the RT. The HR:speed relationship can be obtained either from riding an incremental test in an indoor school or from an unridden lunge test.

Osteocalcin (OC), bone fraction of alkaline phosphatases (BAP) and hydroxyproline (HOP) are markers of bone cell activity. The kinetics of these markers and the analysis of their variations could be related to the osteoarticular status (OAS) of young horses. The growth of Thoroughbreds, French Trotters and Selle Français horses was followed up to 18 months. Blood samples were taken regularly to measure OC, HOP and BAP by standardized techniques. The OAS was evaluated by radiographic examination of the limbs. Based on radiographic findings, two groups of horses were investigated, with no lesions or severely affected. Analysis of variance was used to detect the effects of age and breed, and OAS on parameters. The logarithmic model was used to determine the kinetics of the markers. A rapid decrease in marker concentrations with age and differences between breed was observed. At birth, BAP, OC and HOP concentrations were significantly higher in normal horses (1910 UI l− 1, 192 ng ml− 1 and 35 mg l− 1, respectively) than in horses with severe lesions (1620 UI l− 1, 149 ng ml− 1 and 24 mg l− 1, respectively). During the first 6 months, OC, HOP and BAP remained lower in severely affected horses.

At any level of competition, men run faster than women. Consequently, a male speed advantage is often presumed for other species. This assumption was tested in two animals bred for speed: horses and dogs. Results from Thoroughbred (TB), Standardbred (STB) and Greyhound (GH) races were analysed by ANOVA to compare the speeds of victorious males, neutered males (TB and STB only) and females. Separate analyses were run for shorter (TB: ≤ 1609 m, GH: 503 m) and longer (TB: >1609 m, GH: 603.5 m) TB and GH races. All STB races (trotters and pacers) were 1609 m. In TB races, intact males were 0.7% faster than females at ≤ 1609 m (n = 305; P < 0.01) and 1.4% faster at >1609 m (n = 194; P < 0.01). The speed of neutered males was equivalent to that of females at both distances. Gender accounted for 3.8 and 10.7% of the variance in speed at short and long distances, respectively. In STB pacers, intact males were 1.5% faster than females and gender accounted for 10.1% of the variance in speed (n = 96; P < 0.01). Gender was not a significant predictor of STB trotter (n = 95) or GH speed at 503 m (n = 146) or 603.5 m (n = 23). In conclusion, gender has a significant effect on speed of TBs and STB pacers. Although the effect size is small, it may be significant for racing; in a 7 furlong (1408 m) TB race, the 0.7% difference translates to an advantage of several lengths.

Twelve mature and six 2-year-old Arabian horses were used to determine the effect of dietary long-chain polyunsaturated omega-3 fatty acid supplementation on plasma fatty acids and lameness. Lameness scores and stride lengths were measured on day 0. Horses were striated and pair-matched according to age, gender, stride length and, for mature horses, lameness score, and each horse was fed either a treatment diet containing 5.95 g of stabilized omega-3 fatty acids plus a fat carrier (FA), for a total of 19.4 g fat, or a control diet containing 49 g of corn oil (CO) for 75 days. Horses were exercised 5 d week− 1, and blood samples were drawn and body weights recorded on days 0, 25, 50 and 75. Lameness scores and stride lengths were recorded again on day 75. Total plasma omega-3 fatty acid concentrations were higher on all days in FA horses than in CO horses. Total plasma omega-6 fatty acids increased from days 0 to 25, remained elevated through day 50 and returned to baseline on day 75 in all horses. The ratio of plasma omega-6:omega-3 fatty acids was lower in FA horses. Horses on FA had increased plasma docosahexaenoic acid (DHA) on days 25, 50 and 75. No difference in walk stride length was noted; however, FA horses tended to have a longer trot stride after supplementation when compared with CO horses. No differences were seen in prostaglandin E2 (PGE2) metabolite or tumour necrosis factor-α as measured in blood serum. In summary, supplementing omega-3 fatty acids increases plasma DHA, although there was no overall increase in omega-3 in FA horses. While a trend to increase trot stride length was seen, no differences in lameness scores between treatments were noted.

l-Carnitine supplementation can stimulate erythropoiesis, reduce exercise-induced plasma lactate concentrations and decrease post-exercise muscle damage. Next to horses, Greyhounds represent the premier animal racing species and perform short-duration, very high-intensity exercise that has the potential to incur substantial muscle damage. Under resting and standard racing conditions (5/16 mile), we tested the novel hypotheses that l-carnitine supplementation in Greyhounds would: (1) elevate haematocrit at rest and immediately post-exercise; (2) reduce peak post-exercise plasma lactate; and (3) reduce indices of muscle damage (plasma creatine phosphokinase, CPK and aspartate aminotransferase, AST). Six conditioned Greyhounds (30.1 ± 1.6 kg) underwent a randomized placebo-controlled crossover study to determine the effects of 6 weeks of l-carnitine supplementation (100 mg kg− 1 of body weight/day) at rest and following a maximal speed 5/16 mile race. In accordance with our hypotheses, l-carnitine elevated resting and immediately post-race haematocrit (control, 60.1 ± 1.7, l-carnitine, 63.6 ± 1.7; P < 0.05) and reduced peak post-race plasma CPK and AST concentrations (both P < 0.05). Those dogs with the highest peak post-exercise plasma CPK concentrations under placebo conditions evidenced the greatest reduction with l-carnitine supplementation (r = 0.99, P < 0.01). However, contrary to our hypotheses, l-carnitine did not change peak post-exercise plasma lactate concentrations (control, 27.0 ± 2.1, l-carnitine, 27.7 ± 1.3; P>0.05). We conclude that l-carnitine supplementation increases the potential for oxygen transport and reduces plasma indicators of muscle damage, CPK and AST in racing Greyhounds.

This study examined the changing levels of air pollution at major racetracks in the United States over the past 35 years of Grade I Stakes Thoroughbred racing (n = 675) to evaluate the impact on race performance. Information on winning race times, race speed, temperature, track conditions, ozone and particulate matter was analysed using multiple regression and ANOVA. The 8, 9 and 12 furlong races showed slight improvement in winning times over the 35-year time span. Although there was no significant difference in mean race speed in terms of particulate matter (F4,112 = 0.514, α < 0.05), the few races run under ‘hazardous’ ozone levels, based on the Pollutants Standards Index, were markedly slower (F4,251 = 12.365, α < 0.05). A multiple regression analysis of six additional pollutants in the Californian races revealed no significant effects beyond a slight improvement in the 7 furlong races, with decreased nitrogen dioxide levels. Of the nine racetracks with 9 furlong races, winning times in Pimlico and Belmont were approximately 4 s slower than average. Belmont also had the greatest percentage of races with hazardous ozone conditions. Other correlations between year, pollutants and improvements in race time were of small magnitude and may be an artefact of the sample size rather than of true predictive value. While racing under high levels of certain pollutants appears detrimental to equine athletes, very few races are run under these conditions and other factors such as breeding, training and individual variation may play a more significant role than this study could document with respect to pollution.

Lameness manifests itself during the movement of the animal or at rest, indicating a structural or functional disturbance in one or more limbs. Videographic analyses allow the qualitative and quantitative investigation of the movement, reducing the subjectivity of evaluations. The objective was to record the alterations in the movement of the head and limbs of horses in a lame and non-lame condition. Video cameras (60 Hz) and computational methods were used. Six Purebred Arabian colts were filmed on a treadmill, at walk and at trot, before and after induction of lameness. Lameness was induced in the left forelimb utilizing a transient lameness model. Markers were placed on the zygomatic process of the temporal bone, spiny process of the sixth thoracic and first lumbar vertebrae and on the proximal phalange. Nine strides were analysed. At walk and at trot, the animals demonstrated two vertical head movements per stride, while after induction of lameness only one head movement was observed per stride, where this movement was of greater amplitude. The head was shown to be in a more elevated position when the lame limb made first contact with the treadmill belt, which was not observed in the case of the healthy limbs. Only with trotting did the lame animals manifest a prolonged duration of the stance phase for all limbs. The lame animals had a longer support time, lifted the non-lame limbs and showed a shorter stride. Videographic analyses offered details of the alterations in the movement of horses, which are important in the diagnosis of lameness.

The present study used the physicochemical approach to characterize the changes in acid–base status that occur in Standardbreds after post-exercise electrolyte supplementation. Jugular venous blood was sampled from six conditioned Standardbreds on two separate occasions, at rest and for 24 h following a competitive exercise test (CET) designed to simulate the speed and endurance test of a 3-day event. After the CETs, horses were given water ad libitum and either a hypotonic commercial electrolyte solution, via nasogastric tube followed by a typical hay/grain meal, or a hay/grain meal alone. The electrolyte supplementation resulted in c. 2 mmol l− 1 decreased plasma [TCO2] during the recovery period as compared with control. The primary contributor to the decreased [TCO2] with electrolyte supplementation was a decreased strong ion difference ([SID]), as a result of the non-significant increase in plasma [Cl− ]. Additionally, electrolyte supplementation resulted in faster restoration of hydration status compared with control, as evidenced by faster recovery of plasma [protein] and total weak acid concentration ([Atot]). It is concluded that oral administration of a hypotonic electrolyte solution after prolonged moderate-intensity exercise diminishes the post-exercise alkalosis, and that recovery of hydration status is still incomplete 24 h after exercise when no electrolytes are given. Thus, supplementation with electrolytes according to estimated sweat losses may attenuate post-exercise increases in plasma [TCO2], which is of significant practical interest to the horse racing community, as a testing threshold of greater than 37 mmol l− 1 is used by many racing jurisdictions to determine whether a horse has been administered an alkalinizing agent.

The aim of this study was to determine the heart rate (HR) and work variables of working cattle dogs during actual mustering exercise using a global positioning satellite (GPS) tracking unit‡ with an integrated HR monitor§. The GPS units allowed tracking of seven different Collie and Kelpie working cattle dogs over a total of ten sessions while employed in their usual role of mustering cattle in three locations in Queensland, Australia. Speed, distance and HR data were collected from the dogs during mustering in a variety of working situations. The working dogs covered distances between 13.3 and 30.2 km during mustering sessions ranging from 1 h 59 min to 4 h 24 min at working speeds of up to 43.7 km h− 1. Working temperatures ranged from 29 to 38°C. HR during working exercise ranged between 120 and 237 bpm and was above 180 bpm for 51–68% of the duration of work sessions. There was a positive linear relationship between speed and HR until HRmax (speed 26.0 km h− 1, 233 ± 4.2 bpm), then HR plateaued (R2 = 97.14%, P < 0.001). This study has documented the type of work done by cattle dogs and has shown that GPS devices and HR monitors can be utilized in field conditions to assess the exercise physiology of dogs.

Five previously untrained yearling sled dogs were evaluated for endurance training-induced changes in maximum oxygen consumption (VO2max) and submaximal blood lactate concentrations. Following 3 weeks of light training followed by 4 weeks of moderate training, VO2max increased by 10%, from 180.2 ± 9.9 to 198.7 ± 19.2 ml kg min− 1 (P = 0.046). Light training was not associated with any increase in VO2max. Blood lactate concentrations at the same absolute intensity decreased by 215%, from 9.2 ± 4.7 to 4.3 ± 2.4 mmol l− 1 (P = 0.022). Speeds associated with oxygen consumptions of 70% VO2max increased by 12%, from 4.8 ± 0.4 to 5.4 ± 0.5 m s− 1 (P = 0.008) and speeds associated with VO2max increased by 21%, from 6.7 ± 0.3 to 8.2 ± 0.7 m s− 1 (P = 0.012).

Muscle function depends in part on the interplay between its activity and its length within the stretch-shortening cycle. The longissimus dorsi is a large epaxial muscle running along the thoracic and lumbar regions of the equine back. Due to its anatomical positioning, the longissimus dorsi has the capability of contributing to many functions: developing bending moments in the dorsoventral and lateral (coupled to axial rotation) directions and also providing stiffness to limit motion in these directions. We hypothesize that the exact function of the longissimus dorsi will vary along the back and between gaits as the relation between activity and motion of the back changes. Electromyograms (EMG) were recorded at walk (inclined and level) and trot (on the level) on a treadmill from the longissimus dorsi at muscle segments T14, T16, T18 and L2. Back motion was additionally measured using a fibre-optic goniometer. Co-contractions of the muscle between its left and right sides were quantified using correlation analysis. A greater dominance of unilateral activity was found at more cranial segments and for level walking, suggesting a greater role of the longissimus dorsi in developing lateral bending moments. Timing of the EMG varied between muscle segments relative to the gait cycle, the locomotor condition tested and the flexion–extension cycle of the back. This supports the hypothesis that the function of the longissimus dorsi changes along the back and between gaits.

Reasons for performing the study: In human exercise physiology, the current gold standard for evaluating aerobic capacity is the measurement of oxygen consumption (VO2) and maximal oxygen uptake (VO2max). The evaluation of VO2 in horses is performed in some laboratories equipped with a treadmill but has only been exceptionally reported in field conditions because of the lack of adapted equipment. Objectives: The aim of this study was (1) to assess the feasibility of VO2 measurement on the track using a recently validated portable breath-by-breath gas analyser system adapted to horses (Cosmed K4b2® and Equimask®), (2) to compare these results with those obtained during a treadmill exercise test and (3) to study correlations between VO2 and physiological parameters usually measured in field condition such as heart rate (HR), lactataemia (LA) and the speed at which HR equals 200 beats per minute (bpm) (V200) or LA 4 mmol l− 1 (VLA4). Methods: Five healthy Standardbred trotters in training were submitted to two stepwise incremental exercise tests, one driven on the racetrack and the other on a high-speed treadmill with a 4% incline. Speed (v), HR, ventilatory parameters and VO2 were continuously recorded throughout the duration of the tests and LA was evaluated after each step. Results: All horses completed the test satisfactorily after an initial acclimatization to the mask. There were marked individual differences in ventilatory strategy, and breathing frequency (Rf) at the higher levels of exercise was noticeably low. The VCO2 measurements were incoherent. There were no significant differences between track and treadmill maximal data obtained during the last step [VO2peak (track: 139.9 ± 8.9 ml kg− 1 min− 1; treadmill: 139.9 ± 13.4 ml kg− 1 min− 1), LAmax (track: 6.5 ± 1.6 mmol l− 1; treadmill: 7.3 ± 3.0 mmol l− 1), HRmax (track: 229 ± 6.2 bpm; treadmill: 222 ± 13 bpm)], although the maximal speed required to reach similar workloads was significantly higher on the track (11.9 ± 0.6 m s− 1vs. 9.7 ± 0.4 m s− 1). The correlation between VO2 and HR (r = 0.87; P < 0.001) and VO2 and LA (r = 0.75; P < 0.0001) during both tests was good but no correlation was found between VO2peak and HRmax, LAmax, V200 or VLA4. Conclusions: This is the first report of a practical portable system to measure VO2 and ventilation continuously during high-speed field exercise tests. However, current mask design markedly influences ventilation and could have prohibited the attainment of VO2max. Furthermore, consistent VCO2 measurements should be implemented by the manufacturers. Potential relevance: Continuous breath-by-breath ventilation and VO2 measurements can be recorded in horses in the field at submaximal levels. With necessary adaptations to the system entailed, this study opens new perspectives in the analysis of physiological and metabolic mechanisms of exercise in the equine species in genuine track conditions.