Udry (1996) uses household survey data and finds that the allocation of resources within households is Pareto inefficient, contradicting the main assumption of most collective models of intrahousehold bargaining. He finds that among plots planted with the same crop in the same year, within a given household, those controlled by women produce lower yields than the men’s plots. This paper challenges that finding. Using an alternative nationally representative dataset, I find that only households in regions geographically proximate to those studied by Udry exhibit Pareto inefficient intrahousehold allocations, while the rest of the country reveals no evidence of Pareto inefficiencies. Households in regions experiencing negative rainfall shocks are on average less likely to exhibit Pareto inefficient intrahousehold allocations, and these negative rainfall shocks are correlated with increases in labor resources allocated to the wife’s plots, further confirming that in bad years, households try to avoid losses from Pareto inefficiency.