Purpose: This study evaluates the impact of an outsourced extension service delivered by Lima Rural Development Foundation (Lima) in the Umzimkhulu district of South Africa. The evaluation is conducted at both the household and program levels. Design/methodology/approach: Household impacts were estimated using two-stage regression with instrumental variables to account for selection bias. Gains in household income were extrapolated to estimate the net incremental benefit of the program and its financial return at district level. These analyses were underpinned by a rigorous sampling design. Findings: The results show that the outsourced extension service made a significant and substantial contribution to household crop income and net income. In addition, clients perceived a range of socio-economic benefits such as better diets, improved product quality, and job creation. Analysis of the financial cost and benefit of the extension service at district level suggests an annual net incremental benefit of R5 million (US$0.4 million), and a 95% probability that returns to the service exceeded 20% in 2012/2013. Practical implications: These results suggest that there is good reason for donors to continue funding effective extension services to small farmers in areas of high agricultural potential. Theoretical implications: This is an empirical study. It uses well-established methods to measure the impact but introduces a local economy multiplier to capture some of the indirect effects of agricultural extension. Originality/value: This research is valuable as very few studies worldwide have evaluated the impact of agricultural extension services at both the farm and program levels.