Several countries insert statutory warnings about the adverse effects of smoking in movie and television (TV) content. This study tests the effectiveness of anti-smoking warnings as a prevention-based stimulus for non-smokers in TV programs. Study 1 (a qualitative study) finds that on-screen smoking can induce viewers to take up smoking. Study 2 tested anti-smoking warnings’ efficacy based on viewers’ smoking habits and finds that warnings are most effective in enabling viewers to resist persuasion when they are non-smokers. Study 3, using an eye-tracking device, examines the role of visual attention in the underlying processes. The proposed model compared four warning timings: no warning, warning prior to the content (PW), warning concurrent with the content (CW), and both warnings (both), which lead viewers to pay attention to the on-screen smoking, influencing critical processing (CP) and attitude toward smoking (ASM). Findings indicate that warning timing has a significant indirect effect on ASM, mediated by attention to the warning and on-screen smoking, and CP. There were no direct effects between warning timings and CP or ASM. Additionally, PW and both demonstrated a significantly higher indirect effect on CP and ASM than CW. As compared to CW, PW and both are more effective at countering the persuasion attempt of on-screen smoking. This study contributes to the literature on statutory warning timings, persuasion knowledge models, and healthcare policy. © 2021 Elsevier Inc.