Based on findings from subclinical OCD checking samples (Cuttler & Graf, 2007; 2008), Cuttler and Graf (2009) suggested that checking may develop to compensate for prospective memory failures. The present study provides the first evidence of prospective memory problems in a clinical sample of OCD checkers (n = 26) compared to a group of healthy controls (n = 26). Consistent with earlier work with subclinical samples, the OCD checkers performed more poorly on an event-based, but not a time-based, prospective memory task. However, the OCD checkers did not report more subjective prospective memory failures than controls and their confidence in prospective memory accuracy was higher than that of controls. An explanation of the inconsistent findings with regard to metamemory in clinical and subclinical OCD checking is provided. Further research assessing both subjective memory performance and the strategies understood to support memory is needed to clarify the contribution of prospective memory to OCD checking.