BACKGROUND: Gestational diabetes mellitus (GDM) is a significant cause of adverse perinatal outcomes and major risk factor for type 2 diabetes in mother and child. Although global prevalence is estimated at 14%, the burden in sub-Saharan Africa remains unclear due to limited data and variable diagnostic protocols. This study aimed to generate a robust estimate of GDM prevalence in sub-Saharan Africa using methodologically comparable studies, and to assess subregional variation. METHODS: We systematically searched Embase, MEDLINE, CINAHL, Global Health, African Journals Online, and African Index Medicus from January 1990 to March 2025 for observational studies of pregnant women in sub-Saharan Africa screened for GDM at ≥24 weeks' gestation using an oral glucose tolerance test and internationally recognized criteria. Studies using inconsistent, unclear or incomplete diagnostic protocols or self-reported data were excluded. Quality was assessed using the Joanna Briggs Institute checklist. Prevalence estimates were pooled using random-effects meta-analysis of Freeman-Tukey-transformed proportions. Subgroup analyses were conducted by subregion, and mixed-effects meta-regression examined study-level moderators. FINDINGS: Fifty-nine studies met the inclusion criteria, of which 49 were selected for meta-analysis based on use of comparable diagnostic criteria. Studies represented 16 countries and involved 27,540 participants. The pooled GDM prevalence was 14.0% (95% CI, 11.6 to 16.5; prediction interval 1.9 to 34.3) with substantial heterogeneity (I²=97.1%). Prevalence varied across subregions: Southern Africa 10.2%, Eastern Africa 13.9%, Western Africa 15.1%, and Central Africa 18.0%. Meta-regression showed that small studies (<300 participants), studies using point-of-care testing, and studies conducted before 2016 reported higher prevalence. Subregional differences persisted after adjustment. INTERPRETATION: When comparable diagnostic protocols are applied, GDM prevalence in sub-Saharan Africa matches the global average, challenging perceptions of a lower regional burden. Subregional variability highlights the need for locally representative data. Standardized diagnostic criteria for epidemiological studies would improve comparability and inform targeted public health interventions.