Hepatitis C virus (HCV) genotype 4 non-a/d subtypes, which frequently have NS5A resistance-associated substitutions, are highly prevalent in sub-Saharan Africa. These subtypes, particularly genotype 4r, have been associated with higher rates of failure of treatment regimens containing the NS5A inhibitors ledipasvir or daclatasvir, which are the most accessible direct-acting antivirals in low-income countries. Clinical evidence regarding the efficacy of re-treatment options for these subtypes is limited. We aimed to evaluate the safety and efficacy of sofosbuvir-velpatasvir-voxilaprevir for the treatment of adults in Rwanda with chronic HCV infection, predominantly of genotype 4, and a history of direct-acting antiviral treatment failure. In this single-arm prospective trial, we enrolled adults (aged ≥18 years) with a HCV RNA titre of at least 1000 IU/mL, and a documented history of direct-acting antiviral failure. Patients were assessed for eligibility at a single study site after referral from hospitals with HCV treatment programmes throughout Rwanda, and participants for whom sofosbuvir-ledipasvir treatment had failed in the previous SHARED trial were also included. Participants with decompensated liver disease or hepatitis B virus co-infection were excluded. Participants were treated once daily with an oral fixed-dose combination tablet containing sofosbuvir (400 mg), velpatasvir (100 mg), and voxilaprevir (100 mg) for 12 weeks. The primary endpoint was the proportion of participants with a sustained virological response 12 weeks after completion of treatment (SVR12) in the intention-to-treat population. Viral sequencing of NS3, NS5A, and NS5B genes was done at baseline in all participants and at end of follow-up (week 24) in participants with treatment failure. The study is registered with ClinicalTrials.gov (NCT03888729) and is completed. Between Sept 23, 2019, and Jan 10, 2020, 49 individuals were screened and 40 participants were enrolled. 20 (50%) were female, 20 (50%) were male, median age was 63 years (IQR 56-68), and median HCV viral load was 6·2 log10 IU/mL (5·8-6·5) at baseline. The genotype subtypes identified were 4r (18 [45%] participants), 4k (six [15%]), 4b (five [13%]), 4q (four [10%]), 4l (two [5%]), 4a (one [3%]), 4m (one [3%]), and 3h (one [3%]). One (3%) genotype 4 isolate could not be subtyped, and one (3%) isolate was of unknown genotype. All successfully sequenced isolates (33 [83%]) had at least two NS5A resistance-associated substitutions and 25 (63%) had three or more. 39 (98% [95% CI 87-100]) participants had SVR12. Seven (18%) participants had a total of ten grade 3, 4, or 5 adverse events, including three (8%) cases of hypertension, and one (3%) case each of cataract, diabetes, gastrointestinal bleeding, joint pain, low back pain, vaginal cancer, and sudden death. Four of these events were categorised as serious adverse events resulting in hospitalisation. The one sudden death occurred at home from an unknown cause 4 weeks after the completion of treatment. No serious adverse event was determined to be related to the study drug or resulted in treatment discontinuation. A 12 week course of sofosbuvir-velpatasvir-voxilaprevir is safe and efficacious for the re-treatment of individuals infected with HCV genotype 4 non-a/d subtypes with frequent baseline NS5A resistance-associated substitutions, following failure of previous direct-acting antiviral treatment. Improved affordability and access to sofosbuvir-velpatasvir-voxilaprevir in regions with these subtypes is crucial. Gilead Sciences.