A Pre-experimental design was adapted for the study, 50 younger children were selected by using stratified random method and 10 change agents were selected by using purposive sampling technique based on their academic performance. Change agents knowledge was assess by pretest questionnaire and each change agents was assign 1 topic. The investigator trained them and makes them to do rehearsal and same questionnaire given to determine their adequate level of knowledge on the next day. The change agents were assigned with 5:1 students. The samples were assembled as a small group in a common place and change agents were imparted knowledge regarding prevention of dental caries to the sample for 45 minutes by using PPT presentation. Evaluation of child to child programme approach was assessed by conduction of post test using self administered questionnaire on the 7th days to all the samples. The result shows that there is significant difference between pre test and post knowledge.  The pre test mean was 8.6 and the post test mean was 11.5. Paired t test shows that it is invariably significant at p< 0.01 which was statistically acceptable level of significance. Hence, child to child approach was effective in improving the knowledge of school children.