Background In Indonesia, the infant mortality rate in 2001 was 50 per 1000 live births, with 34.7% due to perinatal death. This perinatal death was associated with low birth weight (LRW) newborn, which was caused by prematurity, infection, birth asphyxia, hypothermia, and inadequate breast feeding. In developing countries, lack of facilities of LBW infant care leads to the utilization of kangaroo method as care to prevent hypothermia in LBW newborn. Objective To evaluate the differences of hypothermia event and duration of birth weight regain in LBW newborns between early kangaroo care (EKC) and conventional care (CC). Methods This was an open label randomized controlled trial. The 1500-2250 g LBW newborns who were born in Sanglah Hospital were randomized to EKC and CC groups. Results Hypothermia events were found more often in CC group than EKC group (RR = 0.645, 90% CI 0.45 to 0.92, P=0.05). This difference was influenced by breast feeding frequency. Duration of birth weight regain in EKC group (median 5 days (SE 0.31., 90% CI 4.49 to 5.51) was shorter than CC group (median 6 days (SE=0.52, 90% CI 5.15 to 6.85), but this difference wasn't statistically significant (P=0.40). Percentage of birth weight decrease, breastfeeding frequency, and hyperbitirubinemia events that needed phototheraphy were associated with the duration of birth weight increase. Conclusion EKC helps to decrease the incidence of hypothermia events, but fails to shorten duration of birth weight increase. Percentage of birth weight decrease, breast-feeding frequency, and hyperbilirunemia events that need photo. theraphy are associated with the duration of birth weight increase in LBW newborn.