To evaluate the existing prospective observational studies on the morality risk among Parkinson's disease (PD) patients and determine the overall risk ratio (RR) of mortality by conducting a meta-analysis and systematic review. Original articles published in English were searched in PubMed and Embase databases prior to March 2013. Only prospective observational studies providing adjusted risk estimates related to PD and future mortality were considered eligible. Pooled adjusted RR and 95% confidence interval (CI) were computed either by fixed-effects models or by random-effects models. Eight studies with 72,833 participants were identified and analysed. In the pooled analyses, patients with PD had a greater risk of all-cause mortality (RR = 2.22; 95% CI: 1.78–2.77). Subgroup analyses based on the design, gender, follow-up duration and sample size showed that a consistent positive association between PD and the mortality risk in each subgroup. However, no statistical significance was found for the baseline age <65 years (RR = 1.42; 95% CI: 0.72–2.77). PD patients with dementia had particularly high mortality risks (RR = 3.78; 95% CI: 2.06–6.92). This meta-analysis indicated that among patients with PD, the all-cause mortality increased by 2.22-fold compared with the general population. PD patients with dementia particularly had higher risks of mortality.