Abstract. The rates of development of the eggs and larvae in utero and the next two developing ovarioles were measured by ovarian dissection on each day of the pregnancy cycle in tsetse, Glossina morsitans, subject to different feeding regimes. Compared with flies fed four times per pregnancy cycle, flies fed three times per cycle showed a lower pupal production rate (70%), the same (zero) adult mortality, a slightly slower growth rate of the larva and second ovariole only from day 8 onwards, but the same growth rate of the first ovariole. Flies fed only twice per pregnancy cycle produced no pupae, suffered 18% adult mortality and showed a significantly slower growth rate of the larva and second ovariole from days 6 and 7 respectively, but still the growth rate of the first ovariole was barely affected. Flies offered food three times or twice per pregnancy cycle engorged fully at every opportunity, but 16.5% of the flies offered food four times per cycle did not feed on every occasion, while 12–22% did not engorge fully on days 3, 5 or 7. In assessing the applicability of these laboratory results to the field situation the following points must be borne in mind: in the laboratory flies take smaller mean blood meals than in the field; during protein production associated with larval growth the proportion of the blood meal lost to transformation and excretory costs is less than during normal lipid metabolism; the balance between the tsetse's known fertility rate and adult and pupal mortality rates reveals that the abortion rate in the field must be extremely low. The high abortion rates usually observed in laboratory colonies, even when flies are offered food dailyl would be quite untenable in the field and indicate that laboratory conditions impose physiological stresses on the flies that are quite different from those in the field. These facts indicate that three field‐sized meals may be sufficient to meet the energy demands of normal larval development in the field.