Two new rotavirus vaccines have recently been licensed in many countries. However, their efficacy has only been shown against certain serotypes commonly circulating in Europe, North America, and Latin America, but thought to be globally important. To assess the potential impact of these vaccines in sub-Saharan Africa, where rotavirus mortality is high, knowledge of prevalent types is essential because an effective rotavirus vaccine is needed to protect against prevailing serotypes in the community. We did two systematic reviews and two meta-analyses of the most recent published data on the burden of rotavirus disease in children aged under 5 years and rotavirus serotypes circulating in countries in sub-Saharan Africa. Eligible studies were selected from PubMed/Medline, Cochrane Library, EmBase, LILACS, Academic Search Premier, Biological Abstracts, ISI Web of Science, and the African Index Medicus. Depending on the heterogeneity, DerSimonian-Laird random-effects or fixed-effects models were used for meta-analyses. Geographical variability in rotavirus burden within countries in sub-Saharan Africa is substantial, and most countries lack information on rotavirus epidemiology. We estimated that annual mortality for this region was 243·3 (95% CI 187·6-301·7) deaths per 100 000 under 5 years (ie, a total of 300 000 children die of rotavirus infection in this region each year). The most common G type detected was G1 (34·9%), followed by G2 (9·1%), and G3 (8·6%). The most common P types detected were P (35·5%) and P (27·5%). Accurate information should be collected from surveillance based on standardised methods in these countries to obtain comparable data on the burden of disease and the circulating strains to assess the potential impact of vaccine introduction.