BACKGROUND: This was a retrospective cohort study undertaken to assess the rate and pattern of dental caries development in 6-year-old school children followed-up for a period of 5 years, and to identify baseline risk factors that were associated with 5 years caries experience in Malaysian children. METHODS: This 5-years retrospective cohort study comprised primary school children initially aged 6 years in 2004. Caries experience of each child was recorded annually using World Health Organization criteria. The rates of dental caries were recorded in prevalence and incidence density of carious lesions from baseline to final examination. Risk assessment was done to assess relative risk for caries after 5 years in children with baseline caries status. Simple and multiple logistic regression analysis were performed to identify significant independent risk factors for caries. RESULTS: The sample consisted of 1830 school children. All components of DMFT showed significant differences between baseline and final examination. Filled teeth (FT) component of the DMFT showed the greatest increases. Results revealed the initial baseline caries level in permanent dentition was a strong predictor for future caries after 5 years (RR=3.78, 95% CI=3.48-4.10, P<0.001). Logistic regression analysis showed significant association between caries occurrence and residence (urban/rural) (OR=1.80, P<0.001). However, it was not significantly associated with gender and ethnicity. The incidence density of caries, affected persons (IDp) observed from baseline and after 5 years was 5.80 persons/100 person-year of observation. The rate of new caries-affected tooth (IDt) in the period from baseline and after 5-years was 0.76 teeth/100 teeth-year of observation. CONCLUSION: The majority of 12-year-old school children (70%) were caries-free and most of the caries were concentrated in only a small proportion (30%) of them. We found that the presence of caries in permanent teeth at the age of 6 years was a strong predictor of future caries development in this population. The strong evidence of early permanent teeth caries at six years old to predict future caries incidence at 12-year-olds, which could be obtained at almost no cost, questions the need for and cost-effectiveness of expensive technology-based commercial caries predictions kits.