The main goal of this study was to investigate the effects of an experimental program designed to develop the phonological awareness (PA) skills of beginning readers in Turkish, an orthographically transparent language. We administered pre-, post-, and follow-up tests to assess the PA skills of 113 first graders and kindergartners in 2 experimental groups (software and in-class face-to-face training) and a control group. Analysis of variance results indicated that all 3 groups improved in terms of overall PA performance. In Grade 1, the children exposed to training, regardless of type, obtained significantly higher posttest means than those in the control group. In kindergarten, the software training group made more immediate progress than the other 2 groups. However, the immediate gains observed at posttest were not maintained at follow-up. We discuss the relevance to the research literature and implications for practice.