Residues of cover crop grasses release nitrogen (N) to subsequent crops, which can contribute to sustainable agricultural management and prevent increases in N-loss-related microorganisms. Moreover, applying N fertilizer to cover crops can enhance the N-use efficiency and yields of subsequent cash crops and tighten the N cycle in the soil. However, the long-term effects of N fertilization of cover crops on soil microbiota and the N cycle in tropical grass-crop no-till systems are unknown. The aim of this study was to evaluate the long-term effects of the timing of N fertilization of cover crops or maize on crop yields, total microbial abundances and N-cycle gene abundances at the time of maize harvest. We carried out a field experiment with two cover crops (palisade grass (Urochloa brizantha) and ruzigrass (U. ruziziensis) fertilized with 120 kg N ha−1 (ammonium sulfate) at one of three times: (i) broadcast over the green cover crops at 35 days before maize seeding (35 DBS), (ii) broadcast over the cover crop straw residues at 1 day before maize seeding (1 DBS), and (iii) as side-dressing at the maize V4 growth stage according to the conventional method (band-applied 0.05 m from the maize row). A control treatment without N application was also carried out for both cover crop species. Except for the control, 40 kg N ha−1 as ammonium sulfate was subsurface band-applied in all treatments 0.05–0.10 m from the maize row at maize seeding, corresponding to 160 kg N ha−1. The total bacterial, archaeal and fungal abundances and abundances of microbial genes encoding enzymes of the N cycle in the soil were quantified by real-time PCR at the maize harvest stage. Overall, maize yield increased significantly in all N fertilizer applications (average 13 Mg ha−1) compared with the control (6 Mg ha−1) over three growing seasons, with maize following palisade grass having the highest yield. The abundances of archaea and fungi in soil were highest under palisade grass that received N at 35 DBS, with values of 4.6 × 106 and 1.7 × 107 gene copies/g of dry soil, respectively. Both cover crop straw production and N release to the soil were positively correlated with the total microbe densities. When ruzigrass was the cover crop, low N enhanced nifH abundance. Archaeal amoA abundance was positively correlated with cover crop biomass and N release regardless of the N treatment and was highest under palisade grass. Bacterial amoA, nirK, and nirS abundances were highest in soil under ruzigrass and were not linked to cover crop biomass mineralization. We conclude that N fertilizer should be applied using the currently recommended method (40 and 120 kg N ha−1 at seeding and side-dressed in maize, respectively) following palisade grass to achieve high maize yield while controlling the level of N loss from tropical soil via nitrification and denitrification.