OBJECTIVES To systematically review available cohort studies and estimate quantitatively the association between occupational exposure to pesticides and Parkinson's disease (PD). METHODS Studies were identified from a MEDLINE search through 30 November 2011 and from the reference lists of identified publications. Relative risk (RR) estimates were extracted from 12 studies published between 1985 and 2011. Meta-rate ratio estimates (mRR) were calculated according to fixed and random-effect meta-analysis models. Meta-analyses were performed on the whole set of data and separate analyses were conducted after stratification for gender, exposure characterisation, PD cases identification, geographic location, reported risk estimator and cohort study design. RESULTS A statistically significant increased risk of PD was observed when all studies were combined (mRR=1.28; 95% confidence interval [CI]: 1.03-1.59) but there was a high heterogeneity and inconsistency among studies. The highest increased risks were observed for studies with the best design, i.e. reporting PD diagnosis confirmed by a neurologist (mRR=2.56; CI: 1.46-4.48; n=4), for cohort studies reporting incidence of PD (mRR=1.95; CI: 1.29-2.97; n=3) as well as for prospective cohorts (mRR=1.39; CI: 1.09-1.78; n=6). A significant increased risk was also seen for banana, sugarcane and pineapple plantation workers (mRR=2.05; CI: 1.23-3.42; n=2). CONCLUSIONS The present study provides some support for the hypothesis that occupational exposure to pesticides increases the risk of PD.