This paper explores factors such as socio-demographics, income and wealth and e-skills that may influence the adoption of the ICTs at the individual level. We examine to what extent these factors contribute to the digital divide between different social groups in Serbia. We use the panel data from the survey “ICT usage in Republic of Serbia” in 2006 and 2007, covering over 3000 individuals/households, to perform a quantitative analysis of the digital divide and ICT adoption. Based on a classification tree and a logistic regression model, a profile of the typical ICT adopter and non-adopter is identified. The empirical results show the following: (i) the digital divide between age group 55-74 and those younger that 55 slightly increased in 2007 in case of regular Internet use; (ii) at the individual level the ICT adoption (use of PC, Internet and mobile phones) in Serbia is mainly influenced by the income and wealth of an individual, its computer and Internet skills and age; (iii) this result is quite robust across the methodological approaches used; and (iv) the classification tree approach is preferable since it gives the same predictive accuracy as the logistic regression with a more parsimonious model. The policy implications of these results are discussed.