Current automated insulin delivery (AID) systems recommend manual insulin delivery prior to meals (hybrid closed-loop [HCL]). Data are needed on the efficacy of an AID system without meal announcement. In this randomized, open-label, parallel-arm trial, we established participants with type 1 diabetes aged 18-70 years on an open-source AID system, with meal announcement, during a 12-week run-in phase, followed by a 12-week trial phase with assignment in a 1:1 ratio to AID without meal announcement or HCL (continued meal announcement). The primary outcome was the percentage of time in the target glucose range of 70-180 mg/dL (3.9-10.0 mmol/L) in the final 14 days of the trial phase, adjusted for the same metric in the final 14 days of the run-in phase. In total, 73 participants underwent randomization (36 to AID without meal announcement and 37 to HCL). Mean (±standard deviation) time in the target range at the end of the run-in and trial phases was 69 ± 11% and 66 ± 8% in the AID without meal announcement and 70 ± 9% and 69 ± 13% in the HCL group (adjusted difference, -2.2 percentage points; 95% confidence interval: -6.2 to 1.7). In adults with type 1 diabetes, use of an open-source AID system without meal announcement demonstrated glycemic efficacy and achieved a similar time in the target glucose range to use of the same system as HCL.