This article examines the impact of non-farm work on household welfare, differentiated by female, for rural households in the Punjab province of Pakistan. We employ an endogenous switching regression approach that accounts for selection bias due to observable and unobservable factors to examine the factors that influence the household’s decision to participate in non-farm work and the impact of participation on household welfare. Given we find no substantial selection bias on unobservable factors; we also use PSM approach to check the robustness of our results from the ESR estimates. Separate estimates are also provided for male and female to address gender heterogeneity. The empirical results reveal that participation in non-farm work significantly increases per head expenditures and reduces household poverty level. This confirms the potential role of non-farm work in improving rural household welfare and poverty alleviation in rural areas of developing countries.