Rapidly expanding population, increasing urbanization, climate change and declining arable land pose a threat to food security in Sub-Saharan Africa (SSA). Efficiency in agricultural production thus becomes an integral aspect. Seed is a fundamental input, however, seed systems in SSA, more so in underutilized crops such as African Indigenous Vegetables (AIV) remain to be constraining. Using household-level data from 445 AIV producing households in Kenya, this study sought to assess the association between seed innovations and performance, as measured by efficiency and income. The study employs bias-corrected Stochastic Meta-Frontier Approach to estimate technical efficiency (TE), allocative efficiency (AE), economic efficiency (EE) and technology gap ratios (TGRs) while accounting for potential technological heterogeneity as well as self-selection bias. We also implement Inverse Probability Weighted Regression Adjustment (IPWRA) for association between seed innovations and income. Results suggest that adopters of seed innovations outperform their non-adopter counterparts on average in both meta-technical, allocative and economic efficiencies and income.