<p>The efficacy of the theory of planned behavior variables was presently examined in predicting exercise stages of change group membership among Greek nonexercising older adults. Data were collected from 399 Greek nonexercising older adults, men and women, aged 65 to 82&#xa0;years belonging either to the precontemplation or the contemplation exercise stage of change. Data collection took place using self-report questionnaires to assess exercise stages of change and the theory of planned behavior variables of attitude toward exercise, subjective norm, perceived behavioral control, and intention to exercise. A series of linear and logistic regression analyses examined the direct and indirect effects of theory of planned behavior variables in predicting group membership in exercise stages of change. Linear regression showed that attitude toward exercise (<i>b</i> = .47, <i>p</i> &lt; .001) and perceived behavioral control (<i>b</i> = .40, <i>p</i> &lt; .001) but not subjective norm (<i>b</i> = .06, <i>p</i> &gt; .05) positively predicted intention to exercise. Subsequent logistic regression analyses showed that among the TPB variables, stages of change group membership was directly predicted only by intention to exercise (odds ratio = 3.66, <i>p</i> &lt; .001). Indirect effects were found for attitude toward exercise and perceived behavioral control but not subjective norm as initial effects of directly predicting group membership in stages of change (attitude: odds ratio = 3.00, <i>p</i> &lt; .001; perceived behavioral control: odds ratio = 1.92, <i>p</i> &lt; .001) became nonsignificant when intention entered the logistic regression. No indirect effects for subjective norm were concluded as subjective norm did not initially predict intention to exercise. Overall, the present findings support the utility of the theory of planned behavior variables in predicting exercise stages of change group membership among Greek nonexercising older adults.</p>