This article reports on a pilot scheme in reception research focusing on the first and so far only closed subtitling model available on Brazilian open television and provided by its most powerful television network, Rede Globo. Experiments with a small sample of deaf and non-deaf viewers from the city of Fortaleza, in the Northeast of Brazil, suggest that the Globo model is not as effective as might be. The guidelines elaborated in the course of this pilot study are envisaged to inform a larger research project which will be developed throughout the country in the hope of arriving at a closed subtitling model that better serves its main target audience.