1. Body size is a central element in current theories of life-history evolution. Models for optimal age at maturity are based on the assumptions that there is a trade-off between development time and adult size and that larger size provides a reproductive advantage.

2. The results of large, replicated experiments with the water strider Gerris buenoi (Heteroptera: Gerridae) contradict both these assumptions. Individual rearings under field conditions showed that there is a negative, not a positive, correlation between development time and adult size. The physiological basis of growth, with stretch-induced moulting, may provide a partial explanation for this correlation.

3. This study examined a number of fitness components for their correlations with female size: lifetime fecundity, reproductive life span, average volume per egg, total volume of eggs laid, and the proportion of eggs hatched. None of these traits was correlated with female size.

4. The data on water striders suggest an alternative scenario for life-history evolution, in which size is not an adaptive trait, but evolves as a correlated response to selection on other traits. This expands the range of possible models, and opens life-history theory to the debate about adaptation and optimality.