Twelve years ago Blatchford and Mortimore's authoritative review of class size research appeared in this journal. They concluded that a major problem with class size research was the lack of detailed studies of complex classroom processes that might mediate class size effects on pupils' learning. This article reviews two UK class size reviews and quantitative, qualitative and mixed method class size research. Evidence from research, and insights from 30 years of classroom-based inquiry, form the basis for the development of theoretical models of relationships between class size, classroom processes and pupils' learning. Recent research evidence from secondary school classrooms calls into question simple one-way relationships between class size and pupils' learning. Politicians are challenged to face up to the complexities involved and to be open to more flexible approaches to reforming the organisation of teaching and learning in schools that go beyond expensive programmes of crude across-the-board class size reductions. Further class size research is recommended that incorporates sophisticated qualitative methods in order to adequately understand and represent the kinds of teacher and pupil expertise involved in promoting and maximising opportunities for high quality learning in different large and small class contexts in primary and secondary schools.