The objective of the study was to determine the chemical composition of the diet, intake and plasma concentrations of glucose, urea and non-esterified fatty acids (NEFA) during pregnancy in goats grazing a thorn scrubland of the semi-arid region of North Mexico. Additionally, monthly changes during pregnancy of plasma concentrations of luteinizing hormone (LH), growth hormone (GH), insulin (INS) and progesterone (P4) were also measured. A herd of 250 goats was bred in August and blood samples were collected from 12 goats at monthly intervals up to December to evaluate plasma glucose, urea, NEFA, LH, GH, INS and P4. Three goats fitted with oesophageal cannulae (38±1·7 kg BW) were used to obtain diet samples during pregnancy. Response variables were analysed by split-plot ANOVA for repeated measures on animals across time. When significant differences were observed among pregnancy months, multiple comparisons were performed using the PROC GLM-PDIFF option of SAS. The response variables organic matter (OM), crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF), lignin, hemicellulose and cellulose differed across pregnancy. Mean values for the whole period were (g/kg DM) 866 (OM), 149 (CP), 524 (NDF), 364 (ADF), 131 (lignin), 160 (hemicellulose), and 233 (cellulose). Mean values for dry matter intake (DMI, kg/day), metabolizable energy intake (MEI, MJ/day) and metabolizable protein intake (MPI, g/day) were 1·2, 8·9 and 74·0, respectively and were different among periods of sampling (P<0·05). Mean plasma glucose concentration was 41±4·3 mg/dl with no significant differences (P>0·05) across pregnancy. Plasma NEFA concentrations increased (P<0·05) at the end of pregnancy (August, 0·396 v. November, 1·27 mM/l). Plasma urea concentrations were higher (P<0·05) in August (26·3 mg/dl) and November (26·2 mg/dl). Serum urea concentrations were related to MPI (r=0·82, P=0·10) and to the undegraded protein absorbed from the duodenum (r=0·79; P=0·07), but not with the rumen NH3-N concentration (r=−0·25; P=0·68). Hormone plasma concentrations differed (P<0·05) across pregnancy. While the largest serum LH concentration (P<0·05) was observed in mid pregnancy (October, 7·8 ng/ml), the lowest value was depicted at the end of gestation (4·8 ng/ml). A similar pattern was observed in plasma P4, with monthly increases from August to November, and a reduction during the last month of gestation (August, 5·6 v. December 2·5 ng/ml). With respect to the metabolic hormones, a negative correlation (−0·23, P=0·05) was observed between GH and INS, with the largest plasma INS levels in mid pregnancy (October 1·9 ng/ml), the lowest INS levels in December, 0·8 ng/ml and the largest serum GH concentration in December, 20·2 ng/ml. Results from the study indicate that DMI, MEI and MPI in grazing goats are useful indicators of the nutritive value of the forage consumed by goats which allowed them to keep an important steroidogenic luteal activity during gestation (P4=5·7±0·82 ng/ml, litter size=1·44 kids). Nonetheless, the observed plasma levels of glucose, urea, NEFA, INS and GH indicate that a schedule of energy supplementation may be outlined to ensure an adequate nutritional status of goats during late pregnancy and early lactation.