Background: The effect of probiotic therapy on bacterial vaginosis (BV) is controversial. We conducted a meta-analysis of the efficacy and safety associated with probiotic treatment for BV. Methods: We searched multiple databases covering up to 1 March 2018. Studies published as blinded randomized controlled trials (RCTs), comparing treatment using probiotic versus active or placebo control in BV patients were included, with at least one-month follow-up. Random effects model and trial sequential analysis (TSA) were applied. Results: Ten studies (n = 2321) were included. Compared with placebo, the probiotics-only therapy resulted in a beneficial outcome both in clinical cure rate at the 30th day (risk ratio, RR = 2.57; 95% confidential interval, 95% CI: 1.96 to 3.37), and Nugent score (mean difference, MD = −2.71; 95% CI: 3.41 to −2.00). This effect decreased but remained significant after eight weeks. Probiotics-post-antibiotics therapy had a decreased effect only for a short term and possibly among studies with a mostly black study population. No extra adverse events were observed. The TSA suggested a larger sample size for effective evaluation of the probiotics as a supplementary remedy. Conclusions: Probiotic regimes are safe and may exhibit a short-term and long-term beneficial effect for BV treatment. The ethnic-specific result for the probiotic used after antibiotics is worthy of further study.