The present study proposed a Behavioral Model of Retirement Satisfaction (BMRS) and tested hypotheses derived from it against predictions made by the more traditional Activity Theory of Aging. Retired (N=16) and employed (N=16) faculty of a state university monitored their activities and mood for 6 days and participated in an interview designed to elicit general information about yearly activities and during which responses to the Life Satisfaction Index-z were obtained. The results provided almost no support for Activity Theory. Only retirees' estimated yearly frequency of formal activities was positively correlated with LSI-Z scores. Partial support was obtained for the BMRS hypothesis that a positive correlation would exist between daily mood level and time spent in positively reinforcing activities (i.e., activities rated as enjoyable). Contrary to predictions, an index of daily reinforcement value which combined daily activity duration and enjoyability information was not correlated significantly with a measure of daily mood (i.e., Profile of Mood States). Participants' self-monitored activity data was grouped into 11 microsystems, as defined by Bronfenbrenner (1977). A variety of correlational analyses were conducted to identify relationships between time spent in each microsystem and level of daily mood. The analyses revealed several relationships worthy of further investigation. Mood was positively correlated with the amount of time retirees spent entertaining friends at home and employees spent in solitary activities at work. Mood was negatively correlated with the amount of time both retired and employed faculty spent interacting with friends away from home; amount of time retirees spent engaging in solitary activities at home; and amount of time employees spent interacting with coworkers at work. The implications of the above relationships and the utility of both the self-monitoring procedure and the systems analysis were discussed. / Ph. D.