To obtain an overall quantitative estimate of the association between pregnancy and gingival inflammation. Medline and EMBASE databases were searched through August 2011. Prospective cohort or cross-sectional studies assessing the effect of pregnancy on gingival inflammation evaluated by the gingival index (GI) and/or bleeding on probing were included. Meta-analyses were performed if possible. Forty-four articles representing 33 studies (14 cohort and 19 cross-sectional) were included. Meta-analyses, performed whenever possible, revealed (1) a significantly lower GI in pregnant women in the first term compared with those in their second or third term of pregnancy; (2) a lower mean GI score in post-partum women compared with women in their second [WMD = 0.143; 95% CI (0.031; 0.255); p = 0.012] or third term [WMD = 0.256; 95% CI (0.151; 0.360); p < 0.001] of pregnancy, when considering cohort studies; (3) Non-pregnant women had lower mean GI values than women in their second or third term of pregnancy. Small changes in plaque levels were reported. Despite the limited number of studies included in the meta-analyses, the present systematic review confirms the existence of a significant increase in GI throughout pregnancy and between pregnant versus post-partum or non-pregnant women, without a concomitant increase in plaque levels. © 2012 John Wiley & Sons A/S.