BACKGROUND: The interactions between antiretroviral therapy (ART) and high-risk human papillomavirus (HPV) and cervical lesions in women living with HIV are poorly understood. We reviewed the association of ART with these outcomes. METHODS: We did a systematic review and meta-analysis by searching MEDLINE and Embase databases for cross-sectional or cohort studies published in English between Jan 1, 1996, and May 6, 2017, which reported the association of ART with prevalence of high-risk HPV or prevalence, incidence, progression, or regression of histological or cytological cervical abnormalities, or incidence of invasive cervcal cancer. Studies were eligible if they reported the association of combination ART or highly active ART use with the following outcomes: high-risk HPV prevalence; squamous intraepithelial lesion (SIL) or cervical intraepithelial neoplasia (CIN) prevalence, incidence, progression, or regression; and invasive cervical cancer incidence among women living with HIV. We did random-effects meta-analyses to estimate summary statistics. We examined heterogeneity with the I statistic. This review is registered on the PROSPERO database at the Centre of Reviews and Dissemination, University of York, York, UK (registration number CRD42016039546). FINDINGS: We identified 31 studies of the association of ART with prevalence of high-risk HPV (6537 women living with HIV) and high grade cervical lesions (HSIL-CIN2+; 9288 women living with HIV). Women living with HIV on ART had lower prevalence of high-risk HPV than did those not on ART (adjusted odds ratio [aOR] 0·83, 95% CI 0·70-0·99; I=51%, adjusted for CD4 cell count and ART duration), and there was some evidence of association with HSIL-CIN2+ (0·65, 0·40-1·06; I=30%). 17 studies reported the association of ART with longitudinal cervical lesion outcomes. ART was associated with a decreased risk of HSIL-CIN2+ incidence among 1830 women living with HIV (0·59, 0·40-0·87; I=0%), SIL progression among 6212 women living with HIV (adjusted hazard ratio [aHR] 0·64, 95% CI 0·54-0·75; I=18%), and increased likelihood of SIL or CIN regression among 5261 women living with HIV (1·54, 1·30-1·82; I=0%). In three studies among 15 846 women living with HIV, ART was associated with a reduction in invasive cervical cancer incidence (crude HR 0·40, 95% CI 0·18-0·87, I=33%). INTERPRETATION: Early ART initiation and sustained adherence is likely to reduce incidence and progression of SIL and CIN and ultimately incidence of invasive cervical cancer. Future cohort studies should aim to confirm this possible effect. FUNDING: UK Medical Research Council.