A classic framework for soil nitrogen (N) cycling, the hole in the pipe (HIP) model, posits a trade‐off in emissions of nitric oxide (NO) and nitrous oxide (N2O) as a function of soil moisture. This has been incorporated into ecosystem models but not tested experimentally and remains an important uncertainty for understanding potential hotspots of reactive N emissions: poorly drained agricultural soils that experience episodically high moisture following intensive fertilization. We incubated soils at moisture ranging from 44% to 100% water‐filled pore space (WFPS). Counter to HIP, we did not observe a consistent trade‐off in NO and N2O emissions at intermediate moisture levels following fertilization, and prefertilization emissions were low. Emissions of N as N2O exceeded NO by 2–200‐fold at all moisture levels and peaked at 73–82% WFPS. Emissions of NO declined with moisture but remained significant even under saturated conditions. Increases in nitrite and reduced iron at high moisture indicated possible NO production from chemodenitrification. Potential nitrification rates were 100–1,000‐fold greater than potential denitrification. Emission factors for fertilizer N ranged from 0.05% to 0.58% (mean = 0.2%) for NO and from 0.4% to 16.9% (mean = 5.3%) for N2O. Our results caution the use of WFPS to predict NO:N2O emission ratios as often employed in ecosystem models. Subsurface N cycling may suppress emissions of NO relative to N2O, and N2O emissions can persist under saturated conditions. Elevated N2O emissions from in‐field wet spots comprising a small landscape extent could potentially address disparities between top‐down and bottom‐up N2O budgets.