Our understanding of latitudinal life history patterns may benefit by jointly considering age and mass at maturity and growth rate. Additional insight may be gained by exploring potential constraints through pushing growth rates to their maximum and scoring physiological cost-related variables. Therefore, we reared animals of a univoltine Spanish and Belgian population and of a semivoltine Swedish population of the damselfly Enallagma cyathigerum (spanning a latitude gradient of ca 2350 km) in a common environment from the eggs until adult emergence and exposed them to a transient starvation period to induce compensatory growth. Besides age and mass at maturity and growth rate we also scored investment in energy storage (i.e. triglycerides) and immune function (i.e. total activity of phenoloxidase). At emergence, body mass was greater in Spain and Sweden and lower in Belgium, suggesting a genetic component for the U-shaped latitudinal pattern that was found also in a previous study based on field-collected adults. The mass difference between univoltine populations can be explained by the shorter development time in the Belgian population, and this despite a higher growth rate, a pattern consistent with undercompensating countergradient variation. In line with the assumed shorter growth seasons, Belgian and Swedish animals showed higher routine growth rates and compensatory growth after transient starvation. Despite a strong link with metabolic rates (as measured by oxygen consumption) populations with higher routine growth rates had no lower fat content and had higher immune function (i.e. immune function decreased from Sweden to Spain), which was unexpected. Rapid compensatory growth did, however, result in a lowered immune function. This may contribute to the absence of perfect compensating countergradient variation in the Belgian population and the lowest routine growth rates in the Spanish population. Our results underscore the importance of integrating key life historical with physiological traits for understanding latitudinal population differentiation.