The present study examined the interrelationships among cognitive and social rigidity, intelligence and personal adjustment in old age, and evaluated a six-week training program teaching social problem-solving skills to the elderly. Ninety-eight subjects, aged 60.9 to 88.0 years, were randomly assigned to the treatment, placebo control, and the waiting-list groups. Subjects were pretested on the TBR Opposites test (cognitive rigidity), the social MEPS (social rigidity), WAIS Vocabulary, and Adjustment of Old Age scales; and post-tested on social rigidity and adjustment The hypothesis that cognitive rigidity is independent of social rigidity was not fully supported. Adjustment to old age was found to be significantly related to social rigidity and, to a lesser degree, to cognitive rigidity. Only treatment subjects showed an overall, significant pre- to post-training improvement on social MEPS scores. The findings suggest that social rigidity is amenable to change.