The potential for feelings of hopelessness and depression in the aged is well documented. Although studies have examined the role of perceived control in ameliorating depression in the institutionalized elderly, no research has actually measured the perceived causal attributions among depressed, hopeless and/or institutionalized elderly populations. To explore the different types of attributional styles used by depressed/nondepressed elderly women (N=50), the old and the reformulated models of learned helplessness were tested. Pretest scales included the Zung Depression Scale, the Gallagher Well-Being Scale, the Rosenberg Self-Esteem Scale, the Rosow and Breslau Health Scale, the Information subtest of the Wechsler Adult Intelligence Scale, and the subject's education level. Subjects were presented with a Thematic Apperception Test card depicting a success/failure situation and were asked to determine the degree to which task difficulty, luck, effort, and ability contributed to the respective outcomes. Data analysis revealed significant interactions on the luck and ability variables, which upheld the reformulated model predictions that depressed subjects would attribute success to an external agent and failure to an internal agent and nondepressed subjects would attribute failure to an external agent and success to an internal agent. The principle implication is that depression may be reversed by replacing maladaptive cognitions with more vital ones, suggesting that treatment for depression in the elderly should continue to be cognitive therapy. (Tables of data analyses are included in the appendices). (NRB)