The present study examines whether employment in the rural non-farm sector (RNFE) has any poverty-reducing and/or vulnerability-reducing effect in Vietnam and India. To take account of sample selection bias associated with RNFE, we have applied treatment-effects model, a variant of Heckman sample selection model. It is found that log per capita consumption or log mean per capita expenditure (MPCE) significantly increased as a result of access to RNFE in Vietnam and India - which is consistent with poverty reducing role of accessing RNFE - with the aggregate effect larger in Vietnam than in India. Access to RNFE significantly reduces vulnerability too in both countries, implying that diversification of household activities into non-farm sector would reduce such risks. When we disaggregate non-farm sector employment by its type, we find that poverty and vulnerability reducing effects are much larger for sales, professionals, and clerks than for unskilled or manual employment in the non-farm sector in both countries. However, as even unskilled or manual non-farm employment significantly reduces poverty and vulnerability in India and poverty in some years in Vietnam, this has considerable policy significance as the rural poor do not have easy access to skilled non-farm employment.