AbstractWe employed field-based studies, with complementary laboratory-based studies, to investigate social and environmental influences on tactical mate-guarding decisions in amphipods (Crustacea). Firstly, we investigated variation in precopulatory mate-guarding duration in Cammarus duebeni celticus in relation to the social structure of natural populations. Variation in population density of up to two orders of magnitude had no effect on precopula duration, whereas guarding durations increased as the sex ratios of the populations became more male biased. That is, males have some ability to assess the probability of other males taking females into precopula and are prepared to guard for longer as this threat of male: male competition increases. A field demonstration of tactical shifts in reproductive behaviour in response to pertinent social conditions is thus provided. Secondly, the 'habitat segregation' hypothesis, which proposes that positive size-assortative pairing in amphipod populations arises due to variation in the use of micro-habitats, was tested in natural field populations and under laboratory conditions in Echinogammarus marinus. This was necessary in order to distinguish any purely environmental determinants of size-assortment from the role of active decisions by males concerning mate choice and male: male competition. The hypothesis was rejected on the grounds that size-assortative pairing arises under both heterogeneous and homogeneous environmental conditions. Further, in both study species, male and female body size were positively correlated with precopula duration. Thus, indirect competition for access to large, fecund females, based on the timing of male entry into precopula, together with direct aggression, provides the explanation for size-assortative pairing in amphipods.