Cycles of freezing and thawing can disrupt soil microorganisms and soil aggregates, resulting in the release of soluble nitrogen at a time when plant roots are largely inactive.  As such, freeze-thaw cycles may have important implications for soil nitrogen losses and overall ecosystem productivity. Nevertheless, many freeze-thaw cycle studies have subjected soils to temperature cycles of unrealistic magnitude and frequency, making it difficult to place their results in an ecological context. We conducted three experiments using mesocosm and field-based approaches with the aim of examining the effects of realistic freeze-thaw cycles on nitrogen leaching losses and availability in soils. Although greater volumes of leachate were collected from mesocosms subjected to freeze-thaw cycles compared to control mesocosms maintained at a constant temperature above freezing, the concentrations of nitrate and ammonium in leachate did not vary significantly among treatments. Variation in the timing of freeze-thaw cycles had no significant effect on nitrate losses from intact mesocosms or on above-ground productivity in the subsequent growing season. However, a pulse of ammonium was recorded in mesocosms subjected to extra thaws during winter. We also used a field-based N tracer experiment to track the fate of nitrogen added at different intervals from late fall through to early spring. Our results emphasize that realistic temperature cycles must be used in freeze-thaw studies to correctly understand the mechanisms that drive nitrogen dynamics over the fall, winter and spring.