In cold temperate humid regions, peak nitrous oxide (N2O) fluxes coincide with spring snowmelt. However, other extreme climatic events in the non-growing season, such as rainfall that causes temporary thawing at the bare soil surface without snow cover, could be responsible for appreciable fluxes of N2O. The objective of this study was to quantify the proportion of ice-trapped N2O and biologically-produced N2O released following a rain-induced thawing event on soils collected in late fall. The lab experiment was done in soil-filled microcosms after the soil surface was frozen at −20 °C for 2 h, which generated a 1–2 cm ice layer at the soil surface, before they were injected with N2, N2O and 15N2O gas at 5 cm depth and then received different amounts of water with customized rainfall simulators (0, 3, 7.5 or 15 mm simulated rainfall h−1) at 4 °C. Gas samples were collected at 2, 6, 12, 24, 48 and 96 h after gas injection. Soil temperature and water content increased faster after rain-induced thawing than after the air-induced thawing event. The release of ice-trapped N2O was not affected by rain-induced thawing because soil pores filled with melted ice-water, which slowed N2O diffusion and reduced N2O to nitrogen gas (N2), based on the production of 15N–N2. However, high simulated rainfall intensity (7.5 and 15 mm h−1) contributed to greater biological N2O emission after 48 h. We conclude that rainfall on a frozen bare soil surface in cold temperate humid regions may cause N loss through harmless N2, rather than the greenhouse gas N2O.