Dysphagia is a potentially harmful feature, also in Parkinson’s disease (PD). As published prevalence rates vary widely, we aimed to estimate the prevalence of oropharyngeal dysphagia in PD in a meta-analysis. We conducted a systematic literature search in February 2011 and two independent reviewers selected the papers. We computed the estimates of the pooled prevalence weighted by sample size. Twelve studies were suitable for calculating prevalence rates. Ten studies provided an estimate based on subjective outcomes, which proved statistically heterogeneous (p < 0.001), with a pooled prevalence estimate with random effect analysis of 35% (95% CI 28-41). Four studies provided an estimate based on objective measurements, which were statistically homogeneous (p = 0.23), with a pooled prevalence estimate of 82% (95% CI 77-87). In controls the pooled subjective prevalence was 9% (95% CI 2-17), while the pooled objective prevalence was 23% (95% CI 13-32). The pooled relative risk was 3.2 for both subjective outcomes (95% CI 2.32-4.41) and objective outcomes (95% CI 2.08-4.98). Clinical heterogeneity between studies was chiefly explained by differences in disease severity. Subjective dysphagia occurs in one third of community-dwelling PD patients. Objectively measureddysphagia rates were much higher, with 4 out of 5 patients being affected. This suggests that dysphagia is common in PD, but patients do not always report swallowing difficulties unless asked. This underreporting calls for a proactive clinical approach to dysphagia, particularly in light of the serious clinical consequences.