Cross-sectional studies have established a relationship between poverty and food insecurity, but little is known about the acute changes within households that lead to changes in food insecurity. This study examined how changes in income, employment status, and receipt of welfare related to change in severity of food insecurity during 1 y among low-income families. In 2005-2007, 501 families living in market and subsidized rental housing were recruited through door-to-door sampling in high-poverty neighborhoods in Toronto. One year later, families were re-interviewed. The final longitudinal analytic sample included 331 families. Within-household change in income, employment, and welfare receipt were examined in relation to change in severity of food insecurity. Severity was denoted by the aggregate raw score on the Household Food Security Survey Module (HFSSM). Analyses were stratified by housing subsidy status owing to differences in characteristics between households. Food insecurity was a persistent problem among families; 68% were food insecure at both interviews. Severity was dynamic, however, as 73.4% answered more or fewer questions affirmatively on the HFFSM between baseline and follow-up. Among market-rent families, a $2000 gain in income during the year and gain of full-time employment were associated with a 0.29 and 1.33 decrease in raw score, respectively (P < 0.01). This study suggests that improvements in income and employment are related to improvements in families' experiences of food insecurity, highlighting the potential for income- and employment-based policy interventions to affect the severity of household food insecurity for low-income families.