Objectives: This Open Randomized Controlled Trial compares effectiveness and safety of KMI and traditional care in LBW infants.Setting: The largest tertiary care hospital in Colombia, with neonatal intensive care and LBW infants follow-up clinic. Patients: Newborn infants (≤ 2000 g.) who survive the neonatal period and are eligible for an in-patient Minimal Care Unit (MCU). 1084 consecutive deliveries were followed and 746 were randomized (382 to KMI and 364 to control). Interventions: KMI infants were discharged as soon as they were randomized, regardless of weight or gestational age. Infants were kept 24 hr/day in an upright position, in skin-to-skin contact and firmly attached to the mother's chest until KMI wasn't tolerated anymore. Control infants were kept in incubators at the MCU until they satisfied usual discharge criteria. Both groups were followed at term and at 3, 6, 9 and 12 months of corrected age. Outcome measures: mortality, infectious episodes, length of hospital stay, growth and development, sequels and proportion of breast feed infants. Results: Baseline variables were evenly distributed, except weight at recruitment (KMI=1678, controls=1713, p=0.045). Risk of dying was lower in the KMI althought not significant (11/339-3.1%- vs. 19/324 -5.5%- RRO.57, 95%CI 0.27-1.17). There were no difference in growth or development indexes nor in length of breast feeding. KMI infants, specially those weighting under 1200 g. at birth spent less time in hospital although the difference was not significant. Number and severity of infections in survivors was similar. A larger proportion of Kangaroo infants was breast feed up to 3 months of corrected age, but the difference disappeared after that age. Conclusions: Results support earlier findings regarding mortality. Growth indexes are almost identical in the two groups, clarifying a previous finding of poor early growth in Kangaroo infants. Kangaroo care is associated with no excess risk of infection. It seems appropriate to recommend the KMI for populations and where health facilities are similar to ours.