This study examines the outcome of a general program to prevent reading difficulties that integrated systematic phonics teaching in an underpriveliged milieu. Seventeen kindergarten teachers applied the program, while nine continued with their regular activities. Two hundred sixty two students, 165 in the experimental group and 97 in the control group, were evaluated on their knowledge of letters and their metaphonological skills. These evaluations were carried out four times, at the beginning and at the end of kindergarten and the first grade. The results show that the development curve for the knowledge of letters was stronger in the experimental group than in the control group, whereas for metaphonological skills, the curves were similar. Moreover, a low, average or high initial level of knowledge of letters modulated the effect of the intervention; the group of students with a low initial level benefited more from the intervention in terms of their knowledge of letters, while for metaphonological skills, the advantages of the program for low and high level students varied depending on their skills. These results suggest that schools would benefit from systematic phonics teaching from the beginning of kindergarten, particularly for students whose knowledge of letters is weak.