Pupils differ in their popularity in class. Research on students' social relations shows that students like to associate with similar peers or in terms of Schaffer (1996, p. 317): "children of like minds seek each other out". This pattern is based on social comparison theory and has become known as homophily (McPherson, Smith-Lovin & Cook, 2001). Homophily can be based on different dimensions, like age, gender, race, looks, class backgrounds or educational attainment, but also on values, interests and beliefs. Especially age and gender are important aspects for schoolchildren up to early adolescence (McPherson, et al., 2001; Lubbers, 2004) while school attainment also seems relevant (Altermatt & Pomerantz, 2003).This study focuses on the criteria pupils have for selecting friends in class. These were grouped in categories as: looks, interests, sportsmanship, school performance, care, social skills, popularity, etc. It seems plausible that pupils differ in the criteria for friendship. If many pupils value sportsmanship very much, a pupil valuing care as a criterion could be less popular. This particular pupil does in that class not meet the criteria for homophily, will not have many friends and is not likely to participate in a subgroup. The criteria pupils claim to use should relate to the social position pupils have in class. This paper addresses gender differences in social position related to differences in selecting criteria for friendship. 120 students in grade 9 (aged 14-15) were asked to name five of their best friends among students of the same grade at the same school. These data are analysed by the use of NEGOPY software (Richards, 1995). The analyses focus on the students' networks in school. In addition to sociometric data we collected data on the criteria the students have for selecting friends among their peers at school, e.g. looks, leisuretime preferances, school performance, caregiving abillities. Each of the existing networks and each of the pupils outside the networks can be characterized by the criteria they use in selecting friends (Pearson & Michell, 2000). The analyses pay attention to the differences between students very much includeed in social networks and stundents less includeed, with special reference to the gender dimension.Datagathering is finished. Data analyses has started. The expected outcome is a better understanding of the differences between boys and girls in social inclusion processes during adolescence.Altermatt, E.R. & Pomerantz, E.M. (2003). The Development of Competance-Related and Motivational Beliefs: An Investigation of Similarity and Influence Among Friends. Journal of Educational Psychology, 95(1), 111-123.Lubbers, M. (2004). The social fabric of the classroom. Peer relations in secondary education. Groningen: ICO. McPherson, M., Smith-Lovin, L. & Cook, J.M. (2001). Birds of a feather: homophily in social networks. Annual Review of Sociology, 27, 415 - 444. Pearson, M. & Michell, L. (2000) Smokerings: Social network analysis of friendship groups, smoking and drug-taking. Drugs: Education, Prevention and Policy. 7 (1), 21-37. Richards, W.D. (1995). Negopy 4.30. Manual and User's Guide. Burnaby: Simon Fraser University. Schaffer, H.R. (1996). Social development. Oxford: Blackwell.The intention is to publish this paper in an international journal .