The present study compared patterns of anticipatory coarticulation for utterances produced in habitual, loud, and slow conditions by 17 individuals with multiple sclerosis (MS), 12 individuals with Parkinson's disease (PD), and 15 healthy controls. Coarticulation was inferred from vowel F2 frequencies and consonant first-moment coefficients. Rate-related changes in coarticulation differed depending on the particular phonetic events in an utterance. In some instances, the slow condition was associated with stronger anticipatory effects, but in other instances the slow condition was associated with weaker anticipatory effects, relative to other speaking conditions. In contrast, coarticulatory patterns for the loud and habitual conditions typically did not differ. Coarticulatory patterns also tended to be similar among speaker groups within each condition. Finally, when acoustic measures of coarticulation differed among speaking conditions, the direction and magnitude of the effect generally were similar for healthy controls, speakers with MS, and speakers with PD. These results are consistent with studies suggesting mostly preserved patterns of coarticulation for speakers with mild to moderate dysarthria, as well as research indicating only subtle coordination deficits for individuals with dysarthria. The finding that increased loudness had a negligible effect on coarticulation also appears to be at odds with the suggestion that increased loudness stimulates orofacial coordination for speakers with dysarthria, although studies including speakers exhibiting coordination impairments at habitual speaking rates would provide a stronger test of this suggestion. Lastly, the fact that speaking condition similarly affected acoustic measures of anticipatory coarticulation for all speaker groups suggests the feasibility of applying theories and models of speech production for neurologically normal talkers to the study of dysarthria.