This article reports the outcome of a meta-analysis of the relation between the frontal lobes and memory as measured by tests of recognition, cued recall, and free recall. We reviewed experiments in which patients with documented, circumscribed frontal pathology were compared with normal control subjects on these three types of tests. Contrary to conventional wisdom, there is strong evidence that frontal damage disrupts performance on all three types of tests, with the greatest impairment in free recall, and the smallest in recognition. (JINS, 1995, 1, 525–536.)