Fertilizer was applied on snow in January 1981 at 100 kg N•ha−1 as [15N]urea, 15NH4NO3, and NH415NO3 to 11-year-old lodgepole pine (Pinuscontorta Dougl. var. latifolia Engelm.) at Spillimacheen in the British Columbia interior and as [15N]urea (200 kg N•ha−1) to 13-year-old Douglas-fir (Pseudotsugamenziesii (Mirb.) Franco) at Green Mountain, a coastal site in British Columbia. Recovery of labelled N after one growing season was determined in soil and biomass at both sites, and it was also monitored during the growing season in snow and soil at Spillimacheen. At Green Mountain, 5.5% of urea N was recovered in tree biomass, 10.8% in understory, and 33.4% in soil organic N (total recovery 49.7%). Leaching may have contributed to N losses at Green Mountain, but was probably not a direct consequence of the application on snow. At Spillimacheen, total recovery of [15N]urea was 93.3%, with 10.1% in tree biomass, 2.4% in understory, and 80.8% in soil. For 15NH4NO3, recoveries were 5.3% in tree biomass, 2.9% in understory, and 87.0% in soil, for a total of 95.2%. For NH415NO3, recoveries were 1.9% in tree biomass, 3.4% in understory, and 39.1% in soil, for a total of 44.4%. At Spillimacheen, the performance of 15NH4NO3 was comparable to that of urea in tree uptake and soil retention. There were large losses with the NH415NO3 source, however, most likely due to leaching and denitrification during snowmelt. For this reason, fertilization with nitrate on snow is not recommended because of nitrate's susceptibility to leaching, but urea and ammonium sources may be applied under these conditions.