A health calendar for a 28-day period was completed hy a random sample of 512 families in a typical upstate New York community. The 71,316 personday descriptions yielded information on life events unaccessible by retrospective interviews, such as everyday medical complaints, well-being, taking of medications and “things going wrong.” Events accessible by interview, such as doctor visits, could be measured more accurately by the health calendar. Relatively rare events like hospitalizations or major stress, such as death in a family, are not efficiently assessed by health diaries. It is argued that a combination of calendar and interview adds new dimensions for the study of health and illness behavior. The health calendar proved to be efficient and reliable for recording a wide range of everyday events. It automatically provides time series data for the study of short-term processes in a family context and, in combination with interview data, promises greater theoretical return from morbidity and utilization surveys. The considerably higher cost involved in the field work and the computer analysis, compared with retrospective interviews only, is in our view justified by the greater quantity of data collected and the larger potentialities for complex analysis.