Most ectotherms follow a pattern of size plasticity known as the temperature-size rule where individuals reared in cold environments are larger at maturation than those reared in warm environments. This pattern seems maladaptive because growth is slower in the cold so it takes longer to reach a large size. However, it may be adaptive if reaching a large size has a greater benefit in a cold than in a warm environment such as when size-dependent mortality or size-dependent fecundity depends on temperature. I present a theoretical model showing how a correlation between temperature and the size–fecundity relationship affects optimal size at maturation. I parameterize the model using data from a freshwater pulmonate snail from the genus Physa. Nine families were reared from hatching in one of three temperature regimes (daytime temperature of 22, 25 or 28 °C, night-time temperature of 22 °C, under a 12L : 12D light cycle). Eight of the nine families followed the temperature-size rule indicating genetic variation for this plasticity. As predicted, the size–fecundity relationship depended upon temperature; fecundity increases steeply with size in the coldest treatment, less steeply in the intermediate treatment, and shows no relationship with size in the warmest treatment. Thus, following the temperature-size rule is adaptive for this species. Although rarely measured under multiple conditions, size–fecundity relationships seem to be sensitive to a number of environmental conditions in addition to temperature including local productivity, competition and predation. If this form of plasticity is as widespread as it appears to be, this model shows that such plasticity has the potential to greatly modify current life-history theory.