The aims of this systematic review (SR) were to evaluate the association between maternal periodontitis and preterm birth (PB) and/or low birth weight (LBW), and the methodological quality of prospective cohort studies conducted for such a purpose. MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials (CENTRAL) databases were searched up to and including October 2010 to identify prospective studies on the association of periodontitis with PB and/or LBW. Search was conducted by two independent reviewers. The methodological quality of the observational studies was assessed using a specially designed methodological tool. Random effects meta-analyses were conducted thoroughly. Search strategy identified 1680 potentially eligible articles, of which 12 prospective studies were included. One cohort study had their data reported in two articles. Of the 11 studies, 10 showed a high methodological quality and one a medium methodological quality. Nine studies (81.8%) found an association between periodontitis and PB and/or LBW. Meta-analysis showed a significant risk of preterm delivery for pregnant women with periodontitis [risk ratio (RR): 1.70 (95% confidence interval (CI): 1.03, 2.81)] and a significant risk for LBW [RR: 2.11 (95% CI: 1.05, 4.23)] or PB/LBW [RR: 3.57 (95% CI: 1.87, 6.84)], as well as a high and unexplained degree of heterogeneity between studies. Although this SR found a consistent association between periodontitis and PB and/or LBW, this finding should be treated with great caution until the sources of heterogeneity can be explained. © 2011 John Wiley & Sons A/S.