Introduction and objective. Bioimpedance analysis (BIA) phase angle (PhA) is an index of the integrity of cells and cellular membranes. The aim of the study was identification of behavioural and anthropometric predictors of PhA in a group of young adults. Materials and method. A cross-sectional observational study of health behaviours, anthropometric indicators and body composition assessed by the BIA method was conducted in a group of Polish young adults (n=92) aged 18-24 (mean-19.33, STD-0.915). Behavioural variables included: level of physical activity, eating behaviours and nutritional knowledge. Body composition was analysed by means of BIA phase-sensitive 8-electrode medical SECA mBCA 525 device. Results. The mean PhA value in the examined cohort was 6.38±0.75 (males-7.22±0.72; females-6.13±0.57). Males also showed higher statistically significant other body composition indices, excluding fat mass. The multiple regression model, including anthropometric variables and gender, which explained the effect of these variables on PhA, occurred to be significant (p<0.0000) and allowed explanation of the 82.49% of PhA variability. PhA was significantly predicted from body mass index (BMI), absolute fat mass, visceral adipose tissue value, skeletal muscle mass value and gender. The regression model, including behavioural predictors and gender, allowed explanation of the lower percentage of PhA variability (42.75%; p<0.0000) and included general intensity of health behaviours, level of nutritional knowledge, and gender. A regression model which would consider simultaneously anthropometric and behavioural variables could not be constructed. Conclusions. In the examined cohort, anthropometric and body composition variables showed a stronger predictive value with respect to PhA, compared to behavioural variables.