To estimate the pooled risk of tobacco smoking for Parkinson's disease (PD) in patients with and without PD family history. We conducted systematic searches of Medline, PsycLIT, Embase, Current contents, Best Evidence, Nisc Mexico Biblioline, previous reviews, examination of cited reference sources and personal contact and discussion with several investigators expert in the field. Studies in all languages were considered. Published observational studies on PD and cigarette smoking stratified by PD family history were reviewed. When two or more papers were based on an identical study, the paper that principally investigated the relationship between PD, smoking stratified by PD family history or the paper that was published last was used. Three case–control studies were carried out between 1996 and 2000, of which one reported risk estimates. The risk of ever smoker in patients with positive PD family history was 0.82 (95% confidence interval 0.44–1.53). There was an obvious protective effect in the pooled estimate in patients with negative PD family history [odds ratio 0.77 (95% confidence interval 0.59–1.01)]. Although our pooled estimates show that smoking is inversely associated with the risk of PD only in patients with negative PD family history, further studies evaluating the interaction between smoking and PD family history are strongly needed.