During 1964–65, 1,646 newly emerged females of Glossina pallidipes Aust., 1,898 of G. palpalis fuscipes Newst. and 2,528 of G. brevipalpis Newst. were marked and released in an area on the north-eastern shore of Lake Victoria in Uganda so that, by examination of those subsequently recaptured, the relation between physiological and calendar age could be determined. In the light of the results, records of physiological age and trypanosome infections in large numbers of wild-caught females of the same three species caught in the same area in 1962–63 were examined, together with published observation on the animals on which they fed, and the effects of age and food-animals on the incidence of infections were investigated.

Physiological age was determined by examination of the reproductive system and was expresses as the exact or approximate number of ovulations that had taken place. Infections with Trypanosoma were classified as brucei-type, vivax-type or congolense-type, and the last two as mature (infective) or immature, according to the sites in which trypanosomes were found in the flies. All brucei-type infections were regarded as mature.

The relation between physiological and calender age indicated that the length of the ovulation cycle was 11 days in G. pallidipes, 11 1/2 in G. brevipalpis and about 15 in G. palpalis fuscipes, but from other evidence that the last figure should have been about 11 days it is considered that ovulation was retarded in the released females of G. palpalis fuscipes.

The incidence of mature vivax- and congolense- type infections rose with age in all three species, beginning for the most part when flies were about a fortnight old. Brucei-type infections were rare and were found only in flies more than about 35 days old; nearly all flies with such infections also had one or both of the other types. The incidence of immature vivax-type infections rose with age up to about 25 days and then levelled off, which suggests that all eventually became mature. Immature congolense-type infections were rare.

G. pallidipes had the highest total infection rate and G. brevipalpis the lowest. The percentage of order flies in the samples was, however, greater in the later than in G. palpalis fuscipes, which suggests that age is not the sole factor determining infection rate. Food-animals appeared to influence it, since the infection rates of the three species were proportional to the percentages of meals obtained from Bovids.

The age at which females were inseminated was lowest in G. palpalis fuscipes, greatest in G. pallidipes and intermediate in G. brevipalpis. Abnormal ovulation cycles were rare in wild-caught flies, and few females that from their age should have been pregnant were found with an empty uterus.