Previously published reports of deficits in linguistically oriented tasks were reviewed for both demented and nondemented Parkinson disease (PD) patients in a discussion of the question of whether PD affects language. Additionally, to illustrate the methodological problems inherent in evaluating linguistic competence, the performance on four linguistically oriented tasks of 12 PD patients who scored 8 or better on the Mental Status Questionnaire was compared to that of 32 age- and education-matched control subjects. These tasks, from which a composite variable representing performance was formed, were oral object description, story retelling, the Peabody Picture Vocabulary Test, and the Similarities subtest of the Wechsler Adult Intelligence Scale (WAIS), a test of verbal associative reasoning. After controlling for the effects of disease severity, a significant effect was observed for the presence of PD. However, when the additional effects of performance on the Block Design subtest of the WAIS were controlled, the PD effect was of borderline significance. These results typify those in the literature, that is, whereas nondemented PD patients can be demonstrated to perform significantly more poorly on linguistically oriented tasks, the performance deficit cannot be conclusively demonstrated to result from a language deficit per se.