A two-microphone method is presented to measure in-duct acoustic properties such as normal impedance, absorption coefficient and transmission loss by using random signal excitation. By measuring the distance between the microphones and the wave spectra the normal impedance, acoustic absorption coefficient and transmission loss can be determined. Bias and random errors are also discussed to estimate the accuracy and precision of estimates of acoustic properties abtained from the acoustic pressure measurements from the two microphones. The analysis shows that the bias errors can be reduced by using a small analysis bandwidth and by measuring the acoustic pressure close to the non-source end of the duct. The random errors, on the other hand, can be minimized by maintaining a high coherence between the acoustic source and the pressure in the duct. Test cases of the method are shown for close tube, single layer glasswool and double layer glasswool.