OBJECTIVES: To develop an EQ-5D-3L Value set for Trinidad and Tobago METHODS: A Bayesian efficient Discrete Choice Experiment (DCE) design was developed using priors that were obtained from a pilot study undertaken in Trinidad and Tobago previously. The design comprised 20 pairs of EQ-5D-3L states. A 5-state Time Trade Off  (TTO) exercise was included in the elicitation protocol. The mean TTO values were used to rescale the coefficients obtained in the DCE on to a 0-100 scale. A representative sample of 300 respondents completed the warm up tasks, TTO and DCE elicitation exercises. RESULTS: Analysis of the data produce an internally valid model with characteristics similar to the model obtained in the pliot (in terms of the relative levels of importance among the 5 dimensions). CONCLUSIONS: The protocol developed for this study can be easily used in the other small developing health systems of the Caribbean which would allow resource allocation decision making to be based on the preferences of the populations of such countries.