Background: Variants in the GBA1 gene, which encodes lysosomal acid glucocerebrosidase, are among the most common genetic risk factors for Parkinson's disease and are associated with faster disease progression. The mechanisms involved are unresolved but might include accumulation of glucosylceramide. Venglustat is a brain-penetrant glucosylceramide synthase inhibitor that, in previous studies, reduced amounts of the glycosphingolipid. We aimed to assess the safety, efficacy, and target engagement of venglustat in people with early-stage Parkinson's disease carrying pathogenic GBA1 variants. Methods: MOVES-PD part 2 was a randomised, double-blinded, placebo-controlled phase 2 study done at 52 centres (academic sites, specialty clinics, and general neurology centres) in 16 countries. Eligible adults aged 18–80 years with Parkinson's disease (Hoehn and Yahr stage ≤2) and one or more GBA1 variants were randomly assigned using an interactive voice–response system (1:1) to 52 weeks of treatment with oral venglustat (15 mg/day) or matching placebo. Investigators, site personnel, participants, and their caregivers were masked to treatment allocation. The primary outcome measure was the change from baseline to 52 weeks in the Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) parts II and III combined score (a higher score indicates greater impairment), and it was analysed in a modified intention-to-treat population (ie, all randomly assigned participants with a baseline and at least one post-baseline measurement during the treatment period). This study was registered with ClinicalTrials.gov (NCT02906020) and is closed to recruitment. Findings: Between Dec 15, 2016, and May 27, 2021, 221 participants were randomly assigned to venglustat (n=110) or placebo (n=111). The least squares mean change in MDS-UPDRS parts II and III combined score was 7·29 (SE 1·36) for venglustat (n=96) and 4·71 (SE 1·27) for placebo (n=105); the absolute difference between groups was 2·58 (95% CI –1·10 to 6·27; p=0·17). The most common treatment-emergent adverse events (TEAEs) were constipation and nausea (both were reported by 23 [21%] of 110 participants in the venglustat group and eight [7%] of 111 participants in the placebo group). Serious TEAEs were reported for 12 (11%) participants in each group. There was one death in the venglustat group owing to an unrelated cardiopulmonary arrest and there were no deaths in the placebo group. Interpretation: In people with GBA1-associated Parkinson's disease in our study, venglustat had a satisfactory safety profile but showed no beneficial treatment effect compared with placebo. These findings indicate that glucosylceramide synthase inhibition with venglustat might not be a viable therapeutic approach for GBA1-associated Parkinson's disease. Funding: Sanofi.