Danish early and late Mesolithic lakeside settlements are examined. Season of occupation is examined through tooth eruption and bone growth of wild boar (Sus scrofa); the early Mesolithic sites were all occupied in summer (including Holmegaard V, previously believed to be a winter settlement), whereas late Mesolithic Ringkloster was occupied in winter and spring. Skeletal element frequency indicates that Ringkloster was a hunting camp and the early Mesolithic sites were not. There were considerable differences between sites of the two periods, despite their very similar locations, because of their different settlement systems. It is therefore argued that Mesolithic settlements cannot be understood by studying individual site location; the settlement system of which each site was a part exerts a crucial influence.