This study assessed the effects of a flexible work schedule (“flextime”) on time allocated to children and spouse by federal workers. Direct behavioral observations of family, home, and work functions were precluded because of the cost involved in observing many people for long periods of time. In order to obtain detailed individual data, participants completed hour‐by‐hour activity logs a mean of twice per week for 35 weeks. Participants received prior training on log completion, initial feedback on the detail of their log entries, and were prompted to complete the forms. Four different procedures assessing reliability indicated a corroboration rate of 80% with other sources. Log data were reliably reduced to nine categories such as “PM time with children” and 37 subcategories such as “time at dinner.” The log data were presented in time‐series form and the use of a quasi‐experimental design showed that participants who altered their work schedule were able to spend more PM time with their families. The log data demonstrated that the capacity exists to assess closely the effects of large‐scale changes at a micro‐behavioral level, but other methods are needed to make complex self‐reporting systems less expensive and more capable of immediate monitoring of t