Caregiver burden is one of the most commonly used variables in caregiving research, both as a predictor and as an outcome. It has been suggested that caregiver burden can be measured in terms of discrete dimensions of well-being and that burden and well-being represent opposite sides of the same coin. The goal of this study was to explore this issue further by comparing commonly used dimensions of burden with parallel dimensions of well-being, both as outcome variables and as predictors of other outcomes, in a sample of caregivers. The findings suggest that burden may tap a unique domain of caregiving outcomes that is not represented by more objective indicators of these effects. Dismissing burden as an unnecessary or redundant construct seems premature.