Complete and competitive markets imply a separation of the consumption (labor supply) and production (labor demand) decisions of the farm household. This paper tests for separation using the observation that in the absence of labor markets, household composition is an important determinant of farm labor use. The test has power when off-farm employment or hiring constraints, or differing efficiencies of family and hired labor lead to demographic variables affecting farm labor demand. An empirical model is developed to test the proposition that household labor demand (farm employment) is independent of family composition. The model is estimated on a household-farm data set from rural Java. Measurement error and endogeneity issues are addressed with instrumental variables techniques. I cannot reject the null hypothesis that farm labor allocation decisions are independent of household structure. The results are robust to different specifications of the labor demand function.