Evolutionary theory predicts that selection will favour sperm traits that maximize fertilization success in local fertilization environments. In externally fertilizing species, osmolality of the fertilization medium is known to play a critical role in activating sperm motility, but there remains limited evidence for adaptive responses to local osmotic environments. In this study, we used a split-sample experimental design and computer-assisted sperm analysis to (i) determine the optimal medium osmolality for sperm activation (% sperm motility and sperm velocity) in male common eastern froglets (Crinia signifera), (ii) test for among-population variation in percentage sperm motility and sperm velocity at various activation-medium osmolalities and (iii) test for among-population covariation between sperm performance and environmental osmolality. Frogs were obtained from nine populations that differed in environmental osmolality, and sperm samples of males from different populations were subjected to a range of activation-medium osmolalities. Percentage sperm motility was optimal between 10 and 50 mOsm kg−1, and sperm velocity was optimal between 10 and 100 mOsm kg−1, indicating that C. signifera has evolved sperm that can function across a broad range of osmolalities. As predicted, there was significant among-population variation in sperm performance. Furthermore, there was a significant interaction between activation-medium osmolality and environmental osmolality, indicating that frogs from populations with higher environmental osmolality produced sperm that performed better at higher osmolalities in vitro. This finding may reflect phenotypic plasticity in sperm functioning, or genetic divergence resulting from spatial variation in the strength of directional selection. Both of these explanations are consistent with evolutionary theory, providing some of the first empirical evidence that local osmotic environments can favour adaptive sperm motility responses in species that use an external mode of fertilization.