Methods for sampling eggs, larvae, and pupae of Aedes spp. in a temporary pond are described. Changes in the area of the pond were mapped by reference to 8 transect lines and correlated with depth at the centre. These pond outlines at different centre depths were also used for the characterization of sampling strata. Samples were taken along the 8 transects only. Soil cores were taken from the dry pond bed, and the larvae that hatched after repeated floodings in the laboratory were used as estimates of the number of viable eggs present in the cores. The temperature of the water used for flooding was adjusted to hatch the ’spring’ species separately from Aedes vexans and Aedes cinereus. Samples were taken in the spring of 1974, in the autumn after the pond had been naturally reflooded, and in the spring of 1975. The estimated number of eggs in the pond decreased with time and the number estimated to have hatched in the summer of 1974 was similar to the estimated number of larvae. The standard errors were about 10% of the mean. Larvae and pupae were sampled with a Belleville sampler and population estimates were obtained for Aedes vexans and Aedes cinereus in August 1974 and for the ’spring’ species, Aedes euedes and Aedes mercurator, in April–May 1975. The efficiency of this program increased when the sample size exceeded 40, when the size of the sample unit was appropriate to the density of larvae, and when sufficient time was allowed for larvae (1 h for instar I) to become trapped. However, standard errors of less than 10% of the mean were obtained only at high mosquito densities.