Sustainable agricultural practices are being promoted across Africa. While literature provides robust evidence on their welfare impacts in isolation, there is limited evidence on how combinations of sustainable agricultural practices contribute to households’ welfare. Due to complementary and substitution effects and cost involved in adopting SAPs, combinations may have impacts that are higher or lower than individual effects. To shed light on this question we employ cross-sectional data from northern Ghana, which was collected from 421 households and 1229 plots. We investigate the adoption and impacts of sustainable agricultural practices (SAPs) on net crop income per acre and consumption expenditure per capita. We employed a maximum simulated likelihood estimation of a Multinomial Endogenous Treatment Effect Model (METEM) to account for observable and unobservable heterogeneity that influences SAP adoption decisions and the outcome variables. Our results reveal that adoption decisions are affected by household and plot level characteristics. We find that adoption of SAPs significantly increase net crop income and consumption expenditure except when soil & water conservation is adopted in isolation. Contrary to some previous studies elsewhere in Africa on this area, we find that SAPs have a stronger effect on income and expenditure when adopted as a package (all together) rather than in isolation or in sub groups.