Continuity of care is a frequently used concept without a commonly accepted definition. Proposed measures reflect different aspects of continuity, the concentration-of-care aspect playing a key role. Almost all previously proposed measures are individual-based. As an alternative we propose a visit-based approach. The visit-based measures are easy to handle technically and, because of immediate interpretations, conceptually. They are more flexible than individual-based measures regarding time perspective and ways of forming subpopulations for different kinds of comparisons. Furthermore, it is possible to derive every individual-based measure hitherto proposed from visit-based measures. When forming population measures from visit-based ones, equal weighting of visits rather than individuals seems natural. We argue that the same weighting scheme is appropriate in many situations when equal weighting of individuals has been used.