OBJECTIVES: To demonstrate a novel way of deriving the incremental cost-effectiveness ratio (ICER) and associated 95% confidence interval (CI) from the cost-effectiveness acceptability curve (CEAC) generated from a family of regressions on net monetary benefit (NMB).  METHODS: Definitions and mathematical properties of the ICER, NMB, and CEAC are explored to construct a technique for deriving 95% CIs around the ICER estimated from the CEAC.   RESULTS: CEA uses the ICER, a measure with statistical issues that preclude easy derivation of confidence intervals.  NMB is defined for any willingness-to-pay (WTP) value as: NMB = (effectiveness X WTP) – cost.  Because NMB is statistically well-behaved, regression analysis can estimate incremental net monetary benefit (INMB) as the parameter estimate associated with treatment.  INMB = (delta effectiveness X WTP) – delta cost.  The CEAC is generated from a family of these regressions where the unique members of the family are identified by unique levels of WTP used to calculate NMB.  The ICER is the point on the CEAC where the probability of being cost-effective is 50%, because at that point INMB is zero and WTP equals delta cost/delta effectiveness; i.e., the ICER.  That point on the CEAC can be identified numerically by simultaneously solving the two equations for INMB from the two regressions that flank estimated INMB of zero.  Knowing estimated INMB and the WTP we have two equations and two unknowns, and we solve for delta effectiveness and delta cost.  We use a similar procedure on the 95% confidence intervals for two estimated INMBs to find the 95% CI for the ICER.  CONCLUSIONS: In the case where we estimate the ICER from a family of regressions on NMB to construct the CEAC we can also find the 95% CI of the ICER.