Obesity is a rapidly spreading endemic in almost every country of the developed world. Due to its severe comorbidities, it represents large risk from the epidemiological point of view. Emphasis is on the early identification of risked individuals, which calls for effective screening methods. In this paper, we examine the possibilities of using laboratory parameters alone for the identification of manifestly obese children – such method can be later used to identify those who are not already obese, but are in risk, i.e. as a screening method. We primarily confine ourselves to univariate analysis, but will also show a multivariate example.