Background and aims Proactive brief cessation advice by a lay counsellor combined with a referral to a smoking cessation service (active referral) is effective in increasing service use and quitting in community smokers. We compared the effect of two modified approaches to referrals on the cessation outcomes in community smokers. Design Three‐arm cluster‐randomized trial. Setting General community in Hong Kong. Participants Daily cigarette smokers (n = 1163; 77.7% male). Interventions Participants were randomized to receive on‐site active referral (OSR, n = 395), where lay counsellors helped participants make appointments with a smoking cessation service of their choice plus tailored reminders; mobile text messaging referral (TMR, n = 385), where participants were encouraged to use a smoking cessation service via text messages; or brief cessation advice only (control, n = 383). Measurements The primary outcome was a self‐reported 7‐day point‐prevalence abstinence at 6 months post‐treatment initiation. Secondary outcomes included 7‐day point‐prevalence abstinence at 3 and 18 months, biochemically validated abstinence, smoking reduction and the use of cessation services at 3, 6 and 18 months. Findings Using intention‐to‐treat analysis, the OSR (17.7%) and TMR (17.1%) groups had significantly higher self‐reported abstinence than the control (12.0%) group at 6 months odds ratio (OR) for OSR versus control = 1.58, 95% confidence interval (CI) = 1.06–2.36; OR for TMR versus control = 1.52, 95% CI = 1.01–2.28; both P < 0.05. The corresponding validated abstinence rates at 6 months were 7.6, 7.8 and 3.9% (OR for TMR versus control = 2.02, 95% CI = 1.07–3.81; OR for TMR versus control = 2.07, 95% CI = 1.10–3.92; both P < 0.05). Self‐reported and validated abstinence were similar at 18 months. OSR groups had higher rates of smoking cessation service use than the control group at all follow‐ups (all P < 0.001). The smoking reduction rates were similar in continuing smokers. Conclusions Simple active referrals (in person or via text messaging) to smoking cessation services increased abstinence rates among smokers in Hong Kong compared with general brief cessation advice. On‐site active referral increased the use of smoking cessation services compared with general brief cessation advice.