The pre-and post-contact mating behaviours of Daphnia pulicaria are investigated by direct observations of vertical distributions, swimming behaviours and male^female interactions. Analysis of vertical distributions in a 1m deep, thermally strati¢ed migration chamber reveals that females were always located in the upper layer of the water column but males exhibited a bimodal depth distribution, in which an individual's depth was a function of body length and water temperature. The observed distributions of males may be the result of several interacting pressures; predation avoidance, life-history optimization, and avoidance of assortative mating. Male swimming behaviour was faster and orthogonal to that of females, which is in agreement with the predictions of encounter-rate maximization models. Video recordings of males and females interacting in a 1-litre vessel showed that males both pursued and contacted other males more often than females. Thus, there was no evidence that Daphnia are able to use water-borne chemical signals to locate and identify potential mates. However, the average duration of male^female contacts (13.8 s) was much longer than those between males (1.6 s), suggesting that males can determine the sex of contacted individuals.Daphnia mating behaviour is signi¢cantly more complex than previously acknowledged. In contrast to the conventional view of Daphnia males swimming more-or-less randomly and mating with any individual encountered, they exhibit behaviours which increase the potential of mating with females while reducing the risk of predation. Several male behaviours, such as`scanning' and the performance of area-restricted spirals upon encounter, are similar to those reported for some copepods and may be common to zooplankton that lack sophisticated chemosensory abilities. The possibility that Daphnia may also be able to assess such important female attributes as species and reproductive status is discussed.