A study of Glossina pallidipes (Austen) was done using various trapping devices. It was observed that groups of females of the same physiological age, the same pregnancy stage, and all replete, entered traps simultaneously. Swarms of teneral females of the same physiological age were also captured simultaneously. These observations led to the conclusion that female tsetse swarm together according to pregnancy stage and/or age groups for the purpose of both looking for hosts and larviposition sites. Laboratory studies on the feeding pattern and behaviour of heavily pregnant female G. pallidipes indicated that only about 5% of such females fed 2 days before larviposition; that is when the breathing lobes of the larva had turned black and it was ready to be larviposited. The rest did not take a meal once they were in the last stage of pregnancy. Field observations of fed females in similar pregnancy stage generally agreed with this, indicating 4.6% as the portion that fed when in that last pregnancy stage. The abortion rate in the field population was estimated to be about 0.1%. The weakness of the method of determining abortion rates in the field is discussed.