We used data from a 9-year mark-recapture study to determine whether demographic factors could explain female-biased sexual size dimorphism in northern water snakes (Nerodia sipedon). Most males reached sexual maturity at 3 years of age, while most females delayed maturity for an additional year. Female survivorship was not significantly lower than that of males, despite the fact that females grow as much as four times faster than males. Among females, survivorship increased until maturity and decreased thereafter, suggesting a survival cost to reproduction. Life-table calculations indicated that the increase in both survival rates and fecundity with body size made 3 years the optimal age for females to reach sexual maturity. However, if females were not large enough at 3 years of age, their best strategy was to mature the following year. Seasonal patterns of mortality suggest that mating imposes a high mortality cost on males. Intermediate-sized males survived slightly but not significantly better than small and large males. This slight survival advantage of intermediate-sized males was not sufficient to explain why males are so much smaller than females. Therefore other selective factors must be responsible for males retaining a small size. A reproductive advantage associated with small size seems the most likely possibility.