The mechanism underlying the association between smoking and other risky behaviours is not clear. In the most of papers smoking is considered as dependent variable. The aim of this study was to recognize tobacco smoking as predictor of more general syndrome measured on Adolescent Risk-Taking Scale (ARTS). The study was conducted in 2010 among 1272 Polish students from 2nd grade of upper-secondary schools. The frequency of tobacco smoking in the last 30 days was measured on scale ranged form 0 (never) do 6 (40 times or more). Risk-taking syndrome was measured by ARTS scale, including six items and being a part of CHIP-AE (Child Health and Illness Profile - Adolescent Version) questionnaire. Standardized 0-100 scale was constructed. All analyses were adjusted by gender. Polish version of ARTS scale has good psychometric properties with one factor structure and reliability at the level of 0.66 (alpha-Cronbach). Results showed that 42.3% of respondents smoke at least once in the past 30 days, without gender-related differences. The mean index of ARTS was equal to 27.8 in boys and 18.5 in girls, respectively. The correlation between results on smoking and ATRS scales was significant (r = 0.300; p < 0.001). Multivariate linear regression model showed that gender and smoking were independent predictor of ARTS score. These two factors explained 14.1% of ARTS variability. Smoking was entered to the model at the first step. Analogous gender specific models indicated that smoking explained 12.3% of ARTS variability in girls and 7.2% in boys. Young people who smoke are at high risk of many adverse effects, including: unsafe driving, sensation seeking and its consequences and delinquent behaviours. Probably they manifest so called negative identity, having lower self-control and lack of skills to refrain from danger situations and behaviors.