This paper identifies some of the barriers to the adoption of car-sharing, termed carpooling in the US, and develops a framework for deriving trusted recommendations based on social network information. The framework is established on a semantic modelling approach putting forward its suitability to resolving adoption barriers while also highlighting the characteristics of trust and social network information that can be exploited. Identification is made of potential vocabularies, ontologies and public social networks which can be used as the basis for deriving direct and indirect trust values in an implementation and will form part of the focus for future work.