College undergraduates (45 males, 88 females) completed measures of life stress, depression, and physical illness on two different occasions, along with measures of hardiness at Time I and perceived social support at Time 2. Structural equation modeling analyses provided only modest support for the hypothesis that hardiness functions prospectively to buffer the effects of stress on depression, and gave no support for the assumed mitigating effects of hardiness on the stress physical illness relationship. Time 2 social support had the largest effect on Time 2 depression, while the best predictors of Time 2 physical illness were Time I physical illness (i.e., the stability coefficient) and Time 2 stress, respectively.