Although Sierra Leone aims to improve the livelihoods of its rice-farming households, the effect of specific policy changes on diverse households remains an open question. This study presents a household model that simulates strategies of production and consumption for rice-farming households in Sierra Leone. It accounts for non-separability between household production and consumption decisions, accounts for resource constraints, and uses a risk-adjusted measure of welfare. The model simulated the effects of three policy scenarios seeking to improve food consumption and calculated production, consumption and socio-economic indicators. The scenarios included: 1) subsidizing rice seed costs, 2) subsidizing palm oil production costs, and 3) subsidizing the conversion of lowlands into land suitable for rice production. The model was applied to four representative rice-farming households in northern Sierra Leone that differed in their land resources and rice consumption. The four household types included: 1) Upland_intensive, 2) Lowland_intensive, 3) Upland_extensive, and 4) Lowland_extensive. Results show that extensive lowland-based households had the lowest rice consumption (54 kg capita−1 year-1) and caloric intake (843 Kcal capita−1 day−1) in comparison with the other household types. Subsidizing rice seed costs in the lowlands, compared to the baseline scenario, improved rice consumption across the households by 15–34%, and farm income by 12–57%. Regardless of the scenario, differences in caloric intake, in comparison to the baseline scenario, were less than differences in rice consumption, thus indicating substitution effects between the different foods (vegetables, cereals, oils, and sweet potatoes). A maximum caloric increase of 16% was achieved by subsidizing palm oil production costs. Subsidizing lowland conversion for rice cultivation improved rice consumption mainly through the subsidy’s positive effect on rice production, which in turn effected self-consumption. Depending on the scenario, the increase in household income was mainly because of an increase in upland rice area, an increase in rice sowing rates, an increase in cash, or a specialization in lowland rice combined with planting a cash crop.