In this paper, we use two different datasets and three different instruments to estimate the impact of a long-established pre-school nursery programme (Hogares Comunitarios) on the nutritional status of beneficiary children. As placement in the programme is endogenous, we use variables related to cost (fee, distance to the nursery) and programme availability (capacity of the programme in the town) as instruments. One of our datasets is representative of very poor individuals living in rural areas of Colombia, while the other focuses in urban areas and include individuals relatively less poor. We find evidence that programme participation increases the height of participating children, with the size of the effect being remarkably consistent across the three instruments and the two datasets, which is informative about the external validity of our estimates. We also pay careful attention to scrutinize the internal validity of the effects that we find.