Abstract.1Blood‐fedAnopheles culicifacieswere collected in a cow‐baited hut, marked with fluorescent powder and released in the same hut. Captured mosquitoes were checked for the presence of marks from the previous day's releases. These checks were carried out on those captured in the release hut and in three similar huts at distances of up to 500 m away.2By dividing the percentage found to be marked in the outlying huts by the percentage marked found on the same day in the release hut, estimates were made of the extent of ‘overlap’ between the populations sampled by the different huts.3Evidence from the rate of build‐up in the percentage marked, from the daily loss rate when a ‘pulse’ of a different colour mark was used, and from the parity rate, gave information on mosquito survival and the interval between marking and recapture. This was notably high in view of earlier field work on the same species in India and of the fact that all houses on the site of the present experiment had been sprayed w