Path analysis was used on survey responses of 1106 persons age 50 or over. Stage one variables were recent retirement, recent widowhood or divorce, age, and sex. Stage two variables included self-rated health, functional health, health trend, and self-rated income adequacy. It was expected that these variables might directly affect morale or indirectly affect it through the stage three variables of total activity space and activity deficit. The stage three variables were also expected to have some direct effect on morale scores. Self-rated health was by far the most important predictor of morale, followed by health trend, activity space, activity deficit, and income adequacy. These five variables accounted for 29% of the variation in morale. (Author/MFD)