Based on case-study data from a small group of Kenya primary schools, this paper first examines local-level variations in the impact of the 2003 Free Primary Education programme. The effects varied widely: in two "high-impact" schools, intakes more than doubled; while in several others, effects were negligible, either because local demand was already fully met, or in two cases because the schools had erected barriers to increased enrolment (the "high-barrier" schools). A second section enquires into secondary-school transition patterns among the graduates from the case-study schools. Although overall transition rates were high, there were wide variations in the type of secondary school entered; and these in turn had massive consequences for the leavers' chances of ultimately qualifying for access to university. The policy implications of these findings are discussed in a final section. (Contains 4 tables and 4 figures.)