OBJECTIVES:  In 2006, the American College of Medical Genetics (ACMG) recommended expanding NBS, relying largely on scoring from a stakeholder survey on 19 attributes of 84 rare conditions. Points were scored according to mean answers from the responders.  Sums of scores resulted in 3 different entry points into an algorithm (EPA) that determined ACMG final screening recommendations.  This research examines one of the survey questions about condition incidence and compares the ACMG use of surveyed opinions versus the actual facts that they also report. METHODS: The report indicated each condition’s mean scores for survey questions.   The incidence question scored 0-100 points. Very rare conditions (<1/100,000) were given no points; conditions with incidence >1/5000 were given 100 points.  Intermediate incidence data received intermediate points (25/50/75).  Alongside these scores was an assessment by ACMG of the incidence.  We rescored the question using the ACMG’s facts on incidence for each condition to see how this would affect total scores and recommendations. RESULTS:  After eliminating conditions with missing data, 78 remained. All had score changes.  We (arbitrarily) defined a positive score change as one where the fact-based score was less than the opinion-based score (OBS – FBS > 0).  Most of the changes were positive (74%). The range was from -37 to +68. The overall mean change was +4 (for positive and negative changes, +12.7 and –20.4). There were two EPA changes, but only one had potential to lead to a change in final recommendation. Thirteen (17%) of the others moved to within 50 points of an EPA change.  CONCLUSIONS: Rescoring this one question had minimal effect on the EPA and therefore on the final ACMG recommendations, however, we only explored one question of the 19.  Similar adjustments to other survey responses could reinforce trends seen here and result in more reclassifications.