Because resources for care of low-birthweight (LBW) infants in developing countries are scarce, the Kangaroo mother method (KMM) was developed. The infant is kept upright in skin-to-skin contact with the mother's breast. Previous studies reported several benefits with the KMM but interpretation of their findings is limited by small size and design weaknesses. We have done a longitudinal, randomised, controlled trial at the Isidro Ayora Maternity Hospital in Quito, Ecuador. Infants with LBW (< 2000 g) who satisfied out-of-risk criteria of tolerance of food and weight stabilisation were randomly assigned to KMM and control (standard incubator care) groups (n = 128 and 147, respectively). During 6 months of follow-up the KMM group had a significantly lower rate than the control group of serious illness (lower-respiratory-tract disorders, apnoea, aspiration, pneumonia, septicaemia, general infections; 7 [5%] vs 27 [18%], p < 0.002), although differences between the groups in less severe morbidity were not significant. There was no significant difference in growth or in the proportion of women breastfeeding, perhaps because the proportion breastfeeding was high in both groups owing to strong promotion. Mortality was the same in both groups; most deaths occurred during the stabilisation period before randomisation. KMM mothers made more unscheduled clinic visits than control mothers but their infants had fewer re-admissions and so the cost of care was lower with the KMM. Since the eligibility criteria excluded nearly 50% of LBW infants from the study, the KMM is not universally applicable to these infants. The benefits might be greater in populations where breastfeeding is not so common.