Reducing nitrogen (N) application represents a strategy for sustainable rice production. The combined effects of deep fertilization (DF) and urease inhibitor (UI) under reduced N input on ammonia (NH3) volatilization, N-use efficiency (NUE) and yield in double-cropping rice remain uncertain. A two-year field trial in Hunan, China, using early-and late-season rice (ESR and LSR), established seven treatments: conventional fertilization (CF, manually broadcast); DF with 30% N reduction; DF with 0.5, 0.75, 1.0, 1.25 and 1.5% NBPT (DF + UI1–DF + UI5, respectively); and a zero-N control (CK). In ESR and LSR, the total NH3 loss rates under CF were 46.9% and 54.2% (2024), and 51.6% and 52.5% (2025), respectively. DF + UI significantly reduced the cumulative NH3 loss rates to 6.9–16.0% (2024) and 9.4–16.4% (2025) for ESR, and to 10.6–17.4% (2024) and 3.6–9.9% (2025) for LSR. DF + UI effectively mitigated NH3 volatilization by lowering surface water ammonium, inhibiting urease activity and delaying urea hydrolysis; it also increased grain yield and apparent N recovery efficiency by 5.6–21.1% and 13.8–33.4 percentage points, respectively. These benefits were attributed to the dual effect of delaying the NH3 volatilization peak by two days while reducing its intensity, which improved synchronization between N supply and rice demand, enhanced total N uptake, promoted N accumulation and increased panicle number and spikelets per panicle. Overall, reducing N input by 30% under DF with 1.25% UI effectively reduced NH3 volatilization while enhancing NUE and rice yield in double-cropping systems.