Intercropping of maize with grain-legume crops predominates on smallholder farms in East Africa. However, the growth and development of crops and their interaction with fertilizer in intercropping systems are not well understood. We sought to answer the questions: (i) what are the effects of fertilizer on the growth and development of maize-pigeonpea and maize-lablab intercropping systems under variable agroecological conditions? (ii) what are the residual effects of two seasons of sole and intercrops of maize, pigeonpea and lablab on a succeeding maize crop? We studied pure stands of maize, long-duration pigeonpea, medium-duration pigeonpea and lablab, and additive intercropping combinations on eight farms in Babati, Tanzania. The intercropping combinations studied were: maize with long-duration pigeonpea, maize with medium-duration pigeonpea and maize with lablab. Three fertilizer levels were applied: no fertilizer; 40 kg P ha−1; and 90 kg N ha−1 + 40 kg P ha−1. The P was applied at planting in the form of triple superphosphate to both maize and legumes, thus in intercrops the 40 kg P ha−1 was shared between maize and legume. The N was spot-applied in the form of urea in three equal splits, only on maize. The trials were implemented for two consecutive cropping seasons (2017/2018 and 2018/2019), without changing treatments and plot locations. To evaluate the residual effect of grain legumes, we planted a sole crop of maize in the third season (2019/2020) in all plots, with no addition of fertilizer. Maize and pigeonpea were sown simultaneously, while lablab was relay-planted one month after maize. Maize dry matter (DM) and grain yield were not significantly affected by the presence of legumes, but were about twice as large in 2017/2018 as in 2018/2019 season, likely due to the considerably higher rainfall in the 2017/2018 season. Legume productivity was more consistent across the two seasons. Legume crops in the intercrops produced significantly less DM and grain yield than in their respective pure stands, but maize yield did not differ significantly between sole and intercrops. The productivity of maize was significantly increased by N fertilization, but the legumes responded positively only to P fertilizer in the 2017/2018 season. Land-equivalent and area-time-equivalent ratios of intercropping systems were consistently greater than one. The DM and grain yield of maize following two seasons of legumes was consistently larger than in plots where maize was grown continuously. The P-fertilizer applied in the preceding seasons had significant residual effects on the yield of the succeeding maize crop. The temporal niche complementarity of pigeonpea and lablab with maize minimized competition in the intercrops, although intercrops are clearly more demanding in labor at planting and harvest. Overall, our results showed the superior performance of maize-legume intercropping over sole maize, both in terms of additional grain yield within a season and the residual effects in the succeeding season. We did not observe biotic constraints in the succeeding maize crop. Maize-legume intercropping with fertilizer application was effective in enhancing the productivity of smallholder cropping systems. [ABSTRACT FROM AUTHOR]