In Experiment 1, it was found that 5-year-oldnew school entrants taught by a syntheticphonics method had better reading, spelling andphonemic awareness than two groups taughtanalytic phonics. The synthetic phonicschildren were the only ones that could read byanalogy, and they also showed better reading ofirregular words and nonwords. For one analyticphonics group the programme was supplemented byphonological awareness training; this led togains in phonemic awareness but not reading orspelling compared with the other analyticphonics group. The synthetic phonics programmewas taught to the analytic phonics groups aftertheir initial programmes had been completed andpost-tested. The group that had hadphonological awareness training did not performbetter than the other two groups when tested 15months later; this was also the case when thesame comparison was made for the the subset ofchildren that had started school with weakphonological awareness skill. Speed of letterlearning was controlled for in Experiment 2; itwas found that the synthetic phonics groupstill read and spelt better than the analyticphonics group. It was concluded that syntheticphonics was more effective than analyticphonics, and that with the former approach itwas not necessary to carry out supplementarytraining in phonological awareness.