The wintering Great Grey Shrike Lanius excubitor was censused in Poland more than two decades ago (during the winters of 1988/89 and 1989/90). Single censuses were made during each winter on sample plots varying from 4.1 to 35.1 km2 (mean: 16.4 km2). Altogether, data from 404 plots, covering in total more than 6.6 thousands km2 were used for the analysis. Based on this data and environmental information gathered in GIS databases (Corine land cover “CLC1990” database, digital elevation model “GTOPO30” dataset) we modelled habitat- and spatial-related variation in shrike distribution. Birds were recorded on 45% of study plots. The mean density was 4.8 individuals/100 km2. In both seasons density in meadows was roughly two times higher than in arable fields. We modelled spatial distribution of wintering birds using combination of GLM and three-dimensional local regression. Models were cross-validated to check their temporal consistency. Repeatable spatial pattern of population distribution allowed to make predictive distribution maps. As a general rule, wintering shrikes avoid regions with severe, continental climate and prefer lowlands dominated by agriculture. On the basis of our models, population of the Great Grey Shrike wintering in Poland during late 1980s can be roughly estimated at about 7.7 thousands of individuals.