To present the epidemiology and psychological predictors of Internet Addiction (IA) among Greek University students.

This random sample consisted of 3545 students, drawn from 24 higher education institutions (1618/1927 M/F, mean age 20.12 ± 2.4 years). The survey was conducted in the 4 Greek cities: Athens, Preveza, Thessaloniki and Amfissa.

The prevalence rates of IA according to Young's Test in the four cities were: Athens (17.5%), Preveza (15.9%), Thessaloniki (19.3%), Amfissa (16.3%). There was a statistically significant association between location and IA (χ2 = 28.251, df = 3, p < 0.001), gender and IA for males and females (p < 0.001). We performed a t-test with the dependent variable S = sum of scores in the psychological features of loneliness, feeling abandoned, disappointed, aimless in life, and with low self-esteem; the grouping variable was IA. Results showed that the variable S was much lower for the non-addicted than for addicted students (t = 19.329, p < 0.005, df = 3085). Binary logistic regression was performed to assess the impact of psychological characteristics on the likelihood that respondents would be Internet Addicted. The model was statistically significant χ2(10, N = 3085) = 316.3, p < 0.001. The model classified correctly 85.1% of the cases. The strongest predictors of IA were the following: Loneliness (OR = 2.15, 95% CI = 1.67-2.71), feeling abandoned (OR = 1.63, 95% CI = 1.2-2.1), low self-esteem (OR = 1.72, 95% CI = 1.4-2.1) and fear of losing their job (OR = 1.36, 95% CI = 1.01-1.7).

Loneliness, helplessness, low self-esteem and fear of unemployment are the strongest predictors of IA among Greek university students.