This study was undertaken to structure and validate a Multi-Professional Family Support Programme that was collectively developed at a Neonatal Intensive Care Unit (NICU). This is participative research of the action-research category with a qualitative–quantitative approach conducted at a University Hospital in the southeast of Brazil with the participation of their multi-professional staff. The study was done in four interdependent stages, adapted from the method proposed by Thiollent (2011): organisation, structuring, validation, and diffusion. Qualitative data were analysed following Bardin’s (2016) Content Analysis Technique and presented in categories and sub-categories. The evaluation instruments followed the evaluation criteria proposed by Pasquali (2010). In the analysis of quantitative data, Cronbach’s Alpha coefficient was used to verify reliability, Fleiss’ Kappa coefficient for measuring agreement, and the Content Validity Index for relevance and representativeness. The programme was built collectively and is based on the Family-Centred Care model. Cronbach’s Alpha reached values above 0.90, which is excellent reliability. There was varying agreement between substantial/perfect and significant (k = 0.68–1.00; p < 0.001), and all the evaluation criteria were considered relevant and representative (CVI > 90.0%). The programme and its guiding technologies were structured and validated with high levels of reliability, agreement, relevance, and representativeness.