<b><i>Background/Aims:</i></b> The aim of this longitudinal case-control study was to investigate variables associated with caries development from birth to 36 months. <b><i>Methods:</i></b> Children (n = 1,017) who were followed up every 6 months from birth to 36 months were grouped into those that developed caries by age 30 and 36 months, respectively, and compared with children without caries. <b><i>Results:</i></b> By 30 months (n = 608) there were 24 children (4%) who had caries and an additional 23 developed first caries at 36 months (n = 552), giving a total prevalence of 47 children with caries (9%) at 36 months. Children who showed caries by 30 months were more likely to be mutans streptococci (MS) colonised by 18 months (p = 0.001) compared to those who developed caries at 36 months, and showed the following variables: MS counts of >10⁵ CFU/ml at 12 months (p = 0.005), missing enamel (p = 0.001), sugar in pacifier at 18 months (p = 0.02), child sleeping next to mother at 6, 18 and 24 months (p = 0.001 to p = 0.02), and exposure to household cigarette smoke at 24 months (p = 0.02). Caries at 36 months was associated with pregnancy problems (p = 0.024), mother having dental cavitations (p = 0.001) and MS presence at 36 months (adjusted odds ratio, AOR = 0.1, p = 0.01 for counts <10⁵ CFU/ml). Caries at both 30 and 36 months was associated with MS presence at 18 months (AOR = 6.3, p = 0.005 and AOR = 4.9, p = 0.01). <b><i>Conclusions:</i></b> Children who developed caries by 30 months are colonised by MS at younger ages and with higher MS counts compared with children who develop caries at 36 months.