OBJECTIVES: Congenital Adrenal Hyperplasia (CAH) is one of the core conditions for which all states plus D.C.  require newborn screening.  Yoo and Grosse (2009) (Y/G) performed a CEA using a national incidence figure, indicating a lack of cost-effectiveness for CAH screening (ICER, $292,000/life year gained). However, the incidence of the disease varies among the states, which could affect the cost-effectiveness of the screening. Here we examine variability in that ICER due to variation in state incidence. METHODS: We reconstructed a decision tree from Y/G to reproduce their deterministic ICER for CAH screening. We then ran the model for every state having a record of cases of CAH in the last year we were able to find incidence measures for most states (2006) to estimate ICERs by state. RESULTS: We were able to approximately reproduce the Y/G result for the ICER  ( $293,927 per life year saved vs. $ 292,000 from Y/G, difference < 1%).  After excluding states with missing data (5) and states with zero measured 2006 incidence (13), the ICERs of the remaining 33 states are shown, as expected, to inversely relate to the state incidence. The ICERs ranged from $37,535 (D.C.) to $1,514,018 (Illinois) per life year saved (median, $303,555). Three of the 33 states have ICERs less than $150,000 per life year. CONCLUSIONS: Different incidences of CAH have a clear effect on the Y/G ICERs of CAH newborn screening. In contrast to the general results from Y/G, using a national average incidence, 3 states have ICERs in a range that indicates possible cost-effectiveness in a deterministic model (ICER < $ 150,000), suggesting that a PSA may be an interesting extension to assess the level of uncertainty around the state-level ICER estimates.