In Experiment 1, it was found that 5-year-old new school entrants taught by a synthetic phonics method had better reading, spelling and phonemic awareness than two groups taught analytic phonics. The synthetic phonics children were the only ones that could read by analogy, and they also showed better reading of irregular words and nonwords. For one analytic phonics group the programme was supplemented by phonological awareness training; this led to gains in phonemic awareness but not reading or spelling compared with the other analytic phonics group. The synthetic phonics programme was taught to the analytic phonics groups after their initial programmes had been completed and post-tested. The group that had had phonological awareness training did not perform better than the other two groups when tested 15 months later; this was also the case when the same comparison was made for the subset of children that had started school with weak phonological awareness skill. Speed of letter learning was controlled for in Experiment 2; it was found that the synthetic phonics group still read and spelt better than the analytic phonics group. It was concluded that synthetic phonics was more effective than analytic phonics, and that with the former approach it was not necessary to carry out supplementary training in phonological awareness.