This paper attempts to provide some evidence related to the ‘feminisation of poverty’, specifically a quantitative evidence related to the greater prevalence of poverty and vulnerability among female-headed households (FHHs) than male-headed households (MHHs). This paper looks at household-level data from 5 different countries (Cameroon, Laos, Madagascar, Mauritania, and Tanzania) from WFP’s Comprehensive Food Security and Vulnerability Analyses (CFSVA) to understand whether factors related to rural employment status impact the level of poverty of FHHs compared to MHHs. We use a specific dimension of poverty as a proxy, i.e. food poverty, to analyse gender aspects of poverty and employment using detailed data at the household level. First we attempt to determine whether there is a clear trend of female poverty over male poverty using stochastic dominance analysis. Preliminary results show that while FHHs are more likely to be food poor related to MHHs in the full sample, this trend becomes less clear when looking only at food poor households. Among food poor households, it seems that it is not unambiguously true that female-headed households are always more poor than male-headed households. In order to better explain this ambiguity, discriminant function analysis was used to identify whether there were statistically significant factors related to rural employment that might explain this phenomenon. The preliminary results of discriminant function analysis show that both types of food poor households face the same obstacles across the 5 countries under study; namely, barriers to access to land, access to productive assets, education, remittances and in some cases over-dependence on subsistence agriculture. These results point to the preliminary conclusion that there are common causes to food poverty that transcend gender differences. This does not mean that MHHs and FHHs follow the same route to this trap.