Understanding why species react differently to changing environments requires detailed understanding of the factors that regulate distributions within current ranges These types of investigations require spatially and temporally explicit examination of the scales at which a species responds to environmental factors To obtain this type of information, we analyzed the relationship between ice cover and abundance, distribution, and spacing of belted kingfishers Ceryle alcyon during the autumn/winters of 1993–1994 and 1994–1995 Our results indicate that the effects of ice cover on kingfisher populations are complex and scale dependent At a small spatial scale (0 5 km) the abundance of kingfishers was negatively correlated with ice cover, as we expected At a large spatial scale (15 5 km), however, there was no relationship between these variables We argue this contradiction m our results is due to differences in the patchiness of ice cover at these scales In a year with average winter temperatures (1993–1994) the spatial and temporal autocorrelation of ice cover and kingfisher distribution were similar In a warm year (1994–1995), however, we found no such similarity Neighbor distances were shorter during the cold year than during the warm year and shorter during censuses with extensive ice cover than censuses with sparse ice cover We conclude that ice cover is a key factor regulating the distribution and abundance of belted kingfishers in our study area during cold to average years, but that during warm years other factors are likely to be more important Based on these patterns we think that patterns of ice cover may be important in mediating the response of kingfishers to changes in climate