Food insecurity is closely related to low agricultural productivity, but it goes beyond basic food production. Poverty, inequalities in access to services, food distribution policies and infrastructure networks play a role in influencing people's access to food. In turn, access to food influences health outcomes on one hand, and on the other, the health of the agricultural labour force influences agricultural output. The literature on the empirical investigations of bi-directional linkages between health and agriculture has been growing, but the findings are affected by the critical lack of allowance for the seasonality of agricultural production, and its differential effects on livelihoods, specifically resulting from the seasonal variations in labour, food stocks, prices, wages, income and expenditure. Using data from the 2010-2011 Integrated Household Survey (IHS3), we use a typology of rural households in the Kasungu-Lilongwe Livelihood Zone of Malawi developed through the technique of cluster analysis. These encompass a diversity of livelihood strategies and outcomes, but the majority of households are very poor with few assets to fall back on in case of shocks. They also suffer regular bouts of ill health. The study then adapts a set of non-linear programming models of the farm household types to simulate and investigate the welfare effects of morbidity, through the interaction between losses in labour and cash resources and the resulting production, consumption and time utilisation responses. Overall, our findings reveal an abundance of family labour but with very limited demand for off-farm employment, and hence households are severely cash constrained. As such, the welfare impacts of morbidity operate particularly through cash losses, where households have to make strategic adjustments on their production and consumption decisions. In addition, the models effectively describe differential responses to similar changes in labour and cash resources across the household types.