OBJECTIVES: Economic evaluation of pharmaceuticals and medical devices has been introduced in France in October 2013. Economic evaluations by HAS are based on critical appraisals of manufacturers’ submissions. We reviewed the selection and use of health utility values for cost-utility analyses (CUA) submitted to HAS. METHODS: A review of manufacturers’ CUA submitted to and assessed by HAS by end of May 2015 was undertaken to review the identification and selection of data and the methods used for deriving utility estimates. The methods used were compared with those recommended in the HAS methodological guideline on economic evaluation. RESULTS: Of the 27 submissions, 16 (56%) included utility values from previously published studies, only 4 of which reported a systematic literature review; 10 (37%) from the clinical trials informing the clinical effectiveness; 2 (7%) from a mixture of sources. To describe health states, 21 (78%) submissions used a preference-based generic instrument, mostly the EQ-5D; 5 (19%) used vignettes; 3 used a condition-specific instrument. Several submissions used different types of instrument. In most (85%) submissions, the valuation method was the time-trade off. The valuation perspective was the general population in 23 (85%) submissions; in only 5 (19%) submissions, the valuation set was derived from the French general population. In 14 (52%) submissions, further adjustments to utility values were made, mostly to account for adverse events. CONCLUSIONS: We identified numerous concerns in the selection, valuation and use of utility values, and a frequent lack of clarity in the description of the methods used. Many submissions included utility values that did not meet the HAS reference case for economic evaluation. There is a need to strengthen early dialogues between manufacturers and HTA bodies to generate robust utility data. Economic evaluation of pharmaceuticals in France will follow a learning curve for both the manufacturers and for HAS.