A distinction in temporal performance has been identified between two classes of rhythmic movements: those requiring explicit timing of salient events marking successive cycles, i.e., event timing, and continuous movements in which timing is hypothesized to be emergent. Converging evidence in support of this distinction is reviewed, including neuropsychological studies showing that individuals with cerebellar damage are selectively impaired on tasks requiring event timing (e.g., tapping). Recent behavioral evidence in neurologically healthy individuals suggests that for continuous movements (e.g., circle drawing), the initial cycle is marked by a transformation from event to emergent timing, allowing the participant to match their movement rate to an externally defined cycle duration. We report a new experiment in which individuals with cerebellar ataxia produced rhythmic tapping or circle drawing movements. Participants were either paced by a metronome or unpaced. Ataxics showed a disproportionate increase in temporal variability during tapping compared to circle drawing, although they were more variable than controls on both tasks. However, two predictions of the transformation hypothesis were not confirmed. First, the ataxics did not show a selective impairment on circle drawing during the initial cycles, a phase when we hypothesized event timing would be required to establish the movement rate. Second, the metronome did not increase variability of the performance of the ataxics. Taken together, these results provide further evidence that the integrity of the cerebellum is especially important for event timing, although our attempt to specify the relationship between event and emergent timing was not successful.