The same pneumococcal conjugate vaccines (PCVs) have been used in adults and children in many settings. Differences in the epidemiology of pneumococcal disease between populations necessitates an adult-specific PCV. We aimed to assess the safety, tolerability, and immunogenicity of V116, an investigational 21-valent PCV designed for adults. This randomised, double-blind, active comparator controlled, international phase 3 trial enrolled adults with or without stable chronic medical conditions at 112 clinical sites in 11 countries or territories. Random assignment was performed using a central electronic interactive response technology system. Cohort 1 (≥50 years) was stratified by age (50-64, 65-74, 75-84, and ≥85 years) and randomised 1:1 to receive one intramuscular dose of V116, or the active comparator, PCV20. Cohort 2 (18-49 years) was randomised 2:1 to receive one intramuscular dose of V116 or PCV20. Pneumococcal serotype-specific opsonophagocytic activity (OPA) and IgG responses were measured before (day 1) and after vaccination (day 30). Four primary immunogenicity outcomes were assessed per-protocol. First, in cohort 1, non-inferiority of V116 to PCV20 was tested using serotype-specific OPA geometric mean titres (GMT) ratios for serotypes common to both vaccines; the lower bound of the 95% CI had to be greater than 0·5 for non-inferiority. Second, superiority of V116 to PCV20 was tested for OPA GMT ratios for the serotypes unique to V116; the lower bound of the 95% CI had to be greater than 2·0 for superiority. Third, superiority of V116 to PCV20 was evaluated by the proportions of participants with a four-fold or greater rise from day 1 to day 30 for serotypes unique to V116; the lower bound of the 95% CI of the differences in proportions (V116 - PCV20) had to be greater than 10% for superiority. Finally, in cohort 2, immunobridging was assessed for all 21 serotypes in V116 for adults aged 18-49 years to 50-64 years; the lower bound of the 95% CI for the OPA GMTs had to be greater than 0·5 for non-inferiority. The safety analysis included all randomly assigned participants who received study vaccine. The primary safety outcome was the proportion of participants with solicited injection site and solicited systemic adverse events until day 5 and vaccine-related serious adverse events up to 6 months after vaccination. This trial is registered at ClinicalTrials.gov (NCT05425732). Between July 13, and Nov 22, 2022, 2754 individuals were screened and 2663 participants were randomly assigned. 2656 individuals were vaccinated (1179 in V116 cohort 1; 1177 in PCV20 cohort 1; 200 in V116 cohort 2; and 100 in PCV20 cohort 2). V116 met non-inferiority criteria compared with PCV20 for the ten serotypes common to both vaccines at day 30 in cohort 1 (p<0·0001 for each common serotype). V116 met superiority criteria compared with PCV20 in cohort 1 for ten of the 11 serotypes unique to V116 at day 30 (OPA GMT ratio: p<0·0001 for all unique serotypes except 15C, which was p=0·41; four-fold or greater rise in OPA from day 1-30: p<0·0001 for all serotypes except 15C, which was p=0·67). Immune responses in V116 participants aged 18-49 years were non-inferior compared with V116 participants aged 50-64 years for all V116 serotypes (p<0·0001 for all V116 serotypes). In cohort 1, 685 (58·2%) of participants in V116, and 778 (66·2%) of participants in PCV20 reported one or more adverse event. In cohort 2, 164 (82·0%) participants in V116 and 79 participants (79·0%) in PCV20 reported one or more adverse event. Six deaths were reported, all in cohort 1, none of which were considered vaccine-related (in V116: one due to sepsis, one due to cerebrovascular accident, one due to myocardial infarction, and one due to hepatic cirrhosis and hepatic encephalopathy; in PCV20: one due to cardiac arrest and one due to abdominal abscess). There were no vaccine-related serious adverse events. V116 was non-inferior to PCV20 for the ten serotypes common to both vaccines and superior to PCV20 for all serotypes unique to V116, except for 15C. Immune responses successfully immunobridged between younger and older adults for all serotypes in V116. V116 was generally well tolerated with safety profile similar to PCV20. Merck Sharp & Dohme, subsidiary of Merck & Co, Rahway, NJ, USA (MSD).