This paper shows how to take into account risk aversion when measuring poverty under income variability. An application to British panel data suggests that income and poverty comparisons between the self-employed and other groups of households are sensitive to assumptions on the degree of risk...