The optimal eating window for time-restricted eating (TRE) remains unclear, particularly its impact on visceral adipose tissue (VAT), which is associated with cardiometabolic morbidity and mortality. We investigated the effects of three TRE schedules (8 h windows in the early day, late day and participant-chosen times) combined with usual care (UC, based on education about the Mediterranean diet) versus UC alone over 12 weeks in adults with overweight or obesity. The primary outcome was VAT changes measured by magnetic resonance imaging. A total of 197 participants were randomized to UC (n = 49), early TRE (n = 49), late TRE (n = 52) or self-selected TRE (n = 47). No significant differences were found in VAT changes between early TRE (mean difference (MD): -4%; 95% confidence interval (CI), -12 to 4; P = 0.87), late TRE (MD: -6%; 95% CI, -13 to 2; P = 0.31) and self-selected TRE (MD: -3%; 95% CI, -11 to 5; P ≥ 0.99) compared with UC, nor among the TRE groups (all P ≥ 0.99). No serious adverse events occurred; five participants reported mild adverse events. Adherence was high (85-88%) across TRE groups. These findings suggest that adding TRE, irrespective of eating window timing, offers no additional benefit over a Mediterranean diet alone in reducing VAT. TRE appears to be a safe, well-tolerated and feasible dietary approach for adults with overweight or obesity. ClinicalTrials.gov registration: NCT05310721 .