The idea that strong social relationships can buffer the negative effects of stress on well-being has received much attention in existing literature. However, previous studies have used less than ideal research designs to test this hypothesis, making it difficult to draw firm conclusions regarding the buffering effects of social support. In this study, we examined the buffering hypothesis in the context of reaction and adaptation to widowhood in three large longitudinal datasets. We tested whether social relationships moderated reaction and adaptation to widowhood in samples of people who experienced loss of a spouse from three longitudinal datasets of nationally representative samples from Germany (N = 1,195), Great Britain (N = 562), and Australia (N = 298). We found no evidence that social relationships established before widowhood buffered either reaction or adaptation to the death of one's spouse. Similarly, social relationships that were in place during the first year of widowhood did not help widows and widowers recover from this difficult event. Social relationships acquired prior to widowhood, or those available in early stages of widowhood, do not appear to explain individual differences in adaptation to loss.