Using multivariate time-series techniques, the dynamics of male and female labor supply and earnings in Puerto Rico are examined during a period of rapid economic developemnt, 1953–1978. The temporal analysis suggests that aggregate male and female labor force participation rates are exogenous to male and female earnings. In addition, continous declines in male participation have been partly responsible for increases in both male and female earnings and a narrowing of relative earnings differentials. The study also finds that the advent of the Food Stamp Program in 1975 has resulted in a small reduction in male labor force participation. Moreover, the inclusion of food stamps as an intervention component is quite useful in forecasting the male participation rate. The forecasting accuracy of univariate and multivariate models is assessed with the result that, in every instance, the multivariate transfer functions prove superior. Overall, the results support the incorporation of multiple time-series in econometric studies of developing countries and provide evidence of its usefullness for development planning and policy.