General circulation models (GCMs) are physically-based models of the climate system that have been used to predict future and explain past climatic changes. Palaeoenvironmental data are required to specify key boundary conditions in model experiments, and then to test model performance. There are however fundamental differences in the spatial and temporal scales at which models simulate and palaeoenvironmental data record climate. Strategies for data-model comparisons allow for these differences, for example by comparing generalized patterns and long-term trends shown in climate simulations with continental-scale syntheses of palaeoenvironmental data. The peak of the last interglacial (Stage 5e) provides an opportunity to use such strategies to examine the effects of insolation changes on regional climates at a time when the insolation anomaly was large and the palaeoenvironmental records unaffected by human impact. GCM simulations of last interglacial climates suggest that the relatively simple patterns of change due to direct insolation effects were radically altered by complex indirect effects due to atmospheric circulation changes. Comparison with Mid-Holocene simulations indicates that the extent and location of the areas affected by such circulation changes may be highly sensitive and non-linearly related to the magnitude of the insolation anomaly. Such indications should be tested by detailed comparisons with palaeoclimatic data on a regional basis.