When judged either by educational attainment of adult population or by secondary and tertiary enrollment rates, Albania by 2002 compared very unfavorably to most European countries, including its neighbors. This study examines the determinants of secondary enrollment applying unobserved family effect probit model to data from Living Standards Measurement Survey 2002-2003. The focus of the paper is to investigate the importance of access to school for enrollment. We find that both absence of a secondary school in the community and the distance from the residence location to a secondary school have strong negative effect on enrollment, controlling for family background. In order to alleviate potential endogeneity bias of distance and community characteristics effects, we control for migration history of individuals since 1990. In rural areas, enrollment is impeded also by absence of a pre-school in the community, and by transportation cost from the community to its \"main\" secondary school. Proximity to a university city (as opposed to other urban centers) substantially increases likelihood of secondary enrollment in rural areas. In urban areas, a similar effect has emerged in 2003, plausibly as a response to opening the market for private universities. The above findings suggest that developing tertiary education and child-care system may have positive spillover effects on secondary enrollment.