Background <p>Access to agricultural credit has long been recognized as a critical determinant of farm productivity, particularly for smallholder farmers in developing countries. In Nigeria, however, restricted access to formal credit remains a significant barrier to achieving improved agricultural performance. This study examines the effect of agricultural credit access on tomato yield among smallholder farmers, with particular attention to correcting for potential selection bias associated with credit participation.</p> Methods <p>Primary data were collected using a multistage sampling technique across major tomato-producing areas in Southwestern Nigeria. Descriptive statistics and the Endogenous Treatment Regression (ETR) model were employed to analyze the data. The ETR approach accounts for non-random selection into credit access by first modeling credit participation using a Probit regression, followed by a second-stage linear regression estimating the effect on tomato yield. Key explanatory variables included demographic characteristics, market access, extension services, and cooperative membership.</p> Results <p>The first-stage Probit regression revealed that age, age squared, farming experience, distance to market, access to extension services, and cooperative membership significantly influenced farmers’ likelihood of accessing credit. In the second stage, the ESR estimates showed that gender, age, age squared, farming experience, number of young dependents, market access, extension services, cooperative membership, and credit access significantly affected tomato yield. The Average Treatment Effect (ATE) of credit access on yield was 8.229, while the Average Treatment Effect on the Treated (ATT) was 8.728—both statistically significant at the 1% level, indicating that credit access enhances tomato productivity. Importantly, the robustness checks using Propensity Score Matching (PSM) and Two-Stage Least Squares (2SLS) returned consistent and comparable estimates of treatment effects, reinforcing the internal validity of the ESR findings. This convergence across methods confirms that the positive impact of credit access on yield is not model-dependent but holds under alternative identification strategies.</p> Conclusions <p>Access to agricultural credit significantly enhances tomato productivity among smallholder farmers in Nigeria. Strengthening institutional frameworks and expanding outreach for agricultural credit provision could serve as effective pathways to boost farm yields. Policymakers are encouraged to integrate credit facilitation into broader agricultural development strategies to support smallholder productivity.</p> Clinical trial registration <p>Not applicable.</p>