Recent work on classroom and school effects in Australia has suggested that teacher effects account for up to 50 per cent of variation in mathematics achievement. The study reported here used data from the Third International Mathematics and Science Study (TIMSS) to look at classroom and school differences in mathematics achievement at both primary and secondary school levels. It found that while classroom differences account for about one quarter of the variation in student achievement, little of this was due to teachers. Most of the classroom variation was due to compositional and organisational factors. This has important implications for policy regarding the improvement of mathematics achievement in schools.