Recent empirical findings have cast doubt on the neoclassical model of labor supply. However, estimation issues, and not workers’ behavior, may be responsible for these findings. This paper investigates this possibility by examining the daily labor supply of Florida lobster fishermen. I invariably find that fishermen work more when earnings are temporarily high, behavior that is consistent with a neoclassical model of labor supply. Furthermore, methods that do not control for measurement error and endogeneity of the wage not only produce downward-biased estimates of labor supply elasticities but also generate a spurious negative and significant elasticity of daily hours.