Background: A phase 2 trial of tocilizumab showed preliminary evidence of efficacy in systemic sclerosis. We assessed skin fibrosis and systemic sclerosis-associated interstitial lung disease (SSc-ILD) in a phase 3 trial to investigate the safety and efficacy of tocilizumab, an anti-interleukin-6 receptor antibody, in the treatment of systemic sclerosis. Methods: In this multicentre, randomised, double-blind, placebo-controlled, phase 3 trial, participants were recruited from 75 sites in 20 countries across Europe, North America, Latin America, and Japan. Adults with diffuse cutaneous systemic sclerosis for 60 months or less and a modified Rodnan skin score (mRSS) of 10–35 at screening were randomly assigned (1:1) with a voice-web-response system to receive subcutaneous tocilizumab 162 mg or placebo weekly for 48 weeks, stratified by IL-6 levels; participants and investigators were masked to treatment group. The primary endpoint was the difference in change from baseline to week 48 in mRSS. Percentage of predicted forced vital capacity (FVC% predicted) at week 48, time to treatment failure, and patient-reported and physician-reported outcomes were secondary endpoints. This trial is registered with ClinicalTrials.gov (number NCT02453256) and is closed to accrual. Findings: Between Nov 20, 2015, and Feb 14, 2017, 210 individuals were randomly assigned to receive tocilizumab (n=104) or placebo (n=106). In the intention-to-treat population, least squares mean [LSM] change from baseline to week 48 in mRSS was −6·14 for tocilizumab and −4·41 for placebo (adjusted difference −1·73 [95% CI −3·78 to 0·32]; p=0·10). The shift in distribution of change from baseline in FVC% predicted at week 48 favoured tocilizumab (van Elteren nominal p=0·002 vs placebo), with a difference in LSM of 4·2 (95% CI 2·0–6·4; nominal p=0·0002), as did time to treatment failure (hazard ratio 0·63 [95% CI 0·37–1·06]; nominal p=0·08). Change in LSM from baseline to week 48 in Health Assessment Questionnaire-Disability Index and in patient-global and physician-global visual analogue scale assessments did not differ between tocilizumab and placebo. In the safety set, infections were the most common adverse events (54 [52%] of 104 participants in the tocilizumab group, 53 [50%] of 106 in the placebo group). Serious adverse events were reported in 13 participants treated with tocilizumab and 18 with placebo, primarily infections (three events, eight events) and cardiac events (two events, seven events). Interpretation: The primary skin fibrosis endpoint was not met. Findings for the secondary endpoint of FVC% predicted indicate that tocilizumab might preserve lung function in people with early SSc-ILD and elevated acute-phase reactants. Safety was consistent with the known profile of tocilizumab. Funding: F Hoffmann-La Roche Ltd.