Background: Spinal muscular atrophy type 1 is a motor neuron disorder resulting in death or the need for permanent ventilation by age 2 years. We aimed to evaluate the safety and efficacy of onasemnogene abeparvovec (previously known as AVXS-101), a gene therapy delivering the survival motor neuron gene (SMN), in symptomatic patients (identified through clinical examination) with infantile-onset spinal muscular atrophy. Methods: STR1VE was an open-label, single-arm, single-dose, phase 3 trial done at 12 hospitals and universities in the USA. Eligible patients had to be younger than 6 months and have spinal muscular atrophy with biallelic SMN1 mutations (deletion or point mutations) and one or two copies of SMN2. Patients received a one-time intravenous infusion of onasemnogene abeparvovec (1·1 × 1014 vector genomes per kg) for 30–60 min. During the outpatient follow-up, patients were assessed once per week, beginning at day 7 post-infusion for 4 weeks and then once per month until the end of the study (age 18 months or early termination). Coprimary efficacy outcomes were independent sitting for 30 s or longer (Bayley-III item 26) at the 18 month of age study visit and survival (absence of death or permanent ventilation) at age 14 months. Safety was assessed through evaluation of adverse events, concomitant medication usage, physical examinations, vital sign assessments, cardiac assessments, and laboratory evaluation. Primary efficacy endpoints for the intention-to-treat population were compared with untreated infants aged 6 months or younger (n=23) with spinal muscular atrophy type 1 (biallelic deletion of SMN1 and two copies of SMN2) from the Pediatric Neuromuscular Clinical Research (PNCR) dataset. This trial is registered with ClinicalTrials.gov, NCT03306277 (completed). Findings: From Oct 24, 2017, to Nov 12, 2019, 22 patients with spinal muscular atrophy type 1 were eligible and received onasemnogene abeparvovec. 13 (59%, 97·5% CI 36–100) of 22 patients achieved functional independent sitting for 30 s or longer at the 18 month of age study visit (vs 0 of 23 patients in the untreated PNCR cohort; p<0·0001). 20 patients (91%, 79–100]) survived free from permanent ventilation at age 14 months (vs 6 [26%], 8–44; p<0·0001 in the untreated PNCR cohort). All patients who received onasemnogene abeparvovec had at least one adverse event (most common was pyrexia). The most frequently reported serious adverse events were bronchiolitis, pneumonia, respiratory distress, and respiratory syncytial virus bronchiolitis. Three serious adverse events were related or possibly related to the treatment (two patients had elevated hepatic aminotransferases, and one had hydrocephalus). Interpretation: Results from this multicentre trial build on findings from the phase 1 START study by showing safety and efficacy of commercial grade onasemnogene abeparvovec. Onasemnogene abeparvovec showed statistical superiority and clinically meaningful responses when compared with observations from the PNCR natural history cohort. The favourable benefit–risk profile shown in this study supports the use of onasemnogene abeparvovec for treatment of symptomatic patients with genetic or clinical characteristics predictive of infantile-onset spinal muscular atrophy type 1. Funding: Novartis Gene Therapies.