OBJECTIVES: To comprehensively search for SF-12 utility scorings and for EQ-5d scorings that can be applied to SF-12 data converted to EQ-5d responses using Gray’s (2006) algorithm. To apply quality/dominance criteria to the identified scorings, apply the scorings of good quality to 3 large US data sets that used the SF12, and examine the implications for scoring strategy. METHODS: A Medline search and hand search of selected journals and of reference lists in relevant articles yielded 27 EQ-5d scorings and 18 studies that provided 29 SF12/SF36 scorings. Quality/dominance criteria excluded 16 scorings from 8 studies. Data sets scored included a survey of 21,564 disabled veterans, a national probability sample of 20,048 veterans, and a longitudinal survey with 4,600 responses from 1,547 hospital-admitted burn victims. We averaged subsets of scores, weighting each component score by the number of people polled in the utility weighting exercise. We also scored 5 extreme value cases.  RESULTS: The composite score performed quite well. Its values and distribution were similar to those in Sengupta’s Health Utility Index 3 scoring and Lawrence’s scoring derived from SF-12 domain scores. Much has been made of the Shaw article that purportedly shows that EQ-5d scoring needs to be country specific. That article used a different scoring equation than other EQ-5d scorings. Its scores differ as much from 4 other US scorings as from overseas scorings – and the other US scores are concordant with the overseas scores. The Gray algorithm performed extremely well, especially when supplemented by questions on use of mobility aids. SF-12 scorings are less dispersed and more prone to floor and ceiling effects than EQ-5d scorings. Time-tradeoff EQ-5d scorings have higher variance than visual-analog scorings. CONCLUSIONS: Some scorings are better than others, with Sengupta or Lawrence recommended. Country of origin rarely matters, with some possible exceptions (Hispanic, African American).