The paper uses cross sectional survey data to analyse poverty among the female-headed households
in a South African Township. Due to the multidimensionality of the term poverty, this study had to specifically
adopt an income definition where households falling below their income poverty line were deemed poor. Binary
logistic regression model was used to estimate the determinants of poverty. The dependant variable was the
households’ economic status (poor and non-poor) and a set of socio-economic and demographic variables used as
explanatory variables The results show that the household size, the age of the household head and total household
income significantly explain variations in the likelihood of being poor. Different methods of survival were
collected for each household. Results indicate several survival strategies, with some having long term health
inferences, calling for immediate policy intervention.