BACKGROUND There is an increasing understanding of the importance of biochar‐based fertilizers in agroecosystems. However, no research has evaluated the effects of partial substitution of urea with biochar‐based urea on rice yields and soil microbial communities. We therefore investigated the rice yields, bacterial communities, and gene abundance involved in nitrogen in silty clay and sandy loam soil paddy fields treated with urea (U), total substitution of urea with biochar‐based urea (BSU), partial substitution of urea with biochar‐based urea in basal and tillering fertilizers (BSU1), and partial substitution of urea with biochar‐based urea in panicle fertilizers (BSU2). RESULTS Compared with U, applying biochar‐based urea increased rice yields, with BSU2 having the most notable effect. Principal coordinate analysis revealed that bacterial communities treated with BSU2 in both soils were significantly different from those treated with U and BSU, most probably due to the decrease in pH caused by the decrease in the concentration of ammonium. The relative abundance of Subdivision3_genera_incertae_sedis, Azotobacter, Geobacter, Buchnera, and Terrimonas in silty clay soils and Saccharibacteria_genera_incertae_sedis and Geobacter in sandy loam soils significantly increased when treated with BSU2 and was positively correlated with rice yields, indicating that the improvements in rice yield were associated with changes in bacterial communities. Based upon amoA/narG related to nitrate accumulation and norB/nosZ related to nitrous oxide emissions, BSU2 enabled a lower risk of nitrate leaching and nitrous oxide emissions in both soils, in comparison with the U and BSU treatments. CONCLUSION The BSU2 treatment had a stronger yield‐increasing effect than biochar‐based urea alone and lowered the risk of nitrogen pollution, which is beneficial to the sustainable development of paddy fields. © 2022 Society of Chemical Industry.