This paper presents two methods for determining the experimental uncertainty of normal-incidence acoustic impedance measurements using the twomicrophone method (TMM). The first method is an analytical technique that extends standard first-order uncertainty methods to complex variables. The second method is a numerical technique that perturbs each variable by its estimated uncertainty. The analytical method provides scaling information on the contribution of each input quantity to the total uncertainty but is limited to small input uncertainties. The numerical method provides more accurate uncertainty estimates and is more amenable to complicated data reduction equations but does not provide scaling information. Two material specimens were tested and the data were reduced using the two methods. The first specimen was a thick piece of aluminum used to represent a sound hard boundary, and the second was a ceramic tubular linear material used to represent a typical acoustic linear material. The results of the analytical and numerical uncertainty methods show good agreement for both specimens.