BACKGROUND Acute paediatric poisoning is a common public health concern for both developed and developing countries. The type of agent and underlying cause differ depending on the social, cultural, economic and educational background. The objectives of this study were to identify the incidence and pattern of paediatric poisoning in a rural district in Sri Lanka and establish whether tertiary referral hospital data are a useful surrogate for estimating district level epidemiology of paediatric poisoning. METHODS A subset of epidemiological data were obtained from March 2011 to February 2013 from a randomized controlled trial (SLCTR/2010/008) conducted in 45 hospitals in Kurunegala district. RESULTS The age adjusted annual incidence of all cause of acute poisoning in children aged 1 to 12 years in the study area was 60.4 per 100,000. The incidence of poisoning of younger age group (1 to 6 years; 76 per 100,000) was significantly higher than older age group (7 to 12 years; 41 per 100,000) (p = 0.0001) in Kurunegala district. The annual incidence rate of paediatric admissions due to deliberate self-poisoning is 18 per 100,000 population. This study also established that admission data from primary hospitals provided the most accurate epidemiological information on paediatric poisoning. CONCLUSIONS In rural districts of Sri Lanka, acute paediatric poisoning cases were less frequent and less severe compared to adult poisoning cases (426-446 per 100,000 population). The incidence of poisoning was significantly higher among young children with compared to old children. In this study, deliberate self-poisoning among older children was more frequently seen than in other comparable countries. Because most of the admissions are directed to and managed by primary hospitals, data from referral hospitals alone cannot be used to represent the true incidence of acute poisoning within a district. The data set from all the primary hospitals (n = 44) yielded more accurate poisoning incidence amongst a paediatric population.