Background: Cigarette smoking is a major public health threat to the world with rising concerns during the COVID-19 pandemic. Objectives: To evaluate the effectiveness of a health belief model (HBM)-based educational intervention in improving knowledge, beliefs, selfreported smoking behaviors, and nicotine dependence among university medical students. Method: A randomized intervention was performed using permuted blocks randomization. Students in the intervention group (n=133) received health education intervention based on the HBM constructs and incorporated smoking related COVID-19 risks. Control subjects (n=129) received basic health education including smoking health related risks. The outcomes measured were: students' knowledge, beliefs, self-reported smoking behaviors, and nicotine dependence. Data were collected at baseline and 30 days post-intervention. Results: After the intervention, percentages of students who had high knowledge and belief scores had significantly increased from ˂5% to >45% in the intervention group (p<0.001). Mean nicotine dependence scores and percentages of daily and heavy smokers among students showed more improvement in the intervention group (3.9±1.5, 35.3%, 5.3%) compared to the control group (5±1.8, 48.8% and 12.4% respectively) (p<0.001) at 30 days post-intervention. Perceived COVID-19 risk susceptibility and risk severity significantly predict post-intervention reduction in moderate/heavy smoking (p<0.01 and p<0.001 respectively) and high nicotine dependence (p<0.01) in the intervention group. Conclusions: This intervention significantly improved knowledge, beliefs, self-reported smoking behaviors and nicotine dependence in medical students. Further research is needed for investigating the effectiveness of applying more complex and longer intervention.