This study examined the effect of external microphone reference placement on peak sound pressure level (pSPL) and measurement variability. Nine normal subjects were seated in a double-walled sound suite, 1 m and 0 degrees azimuth from a wall-mounted speaker. Digitized Gaussian noise was presented at 80 dB pSPL with a 500 msec duration and was measured through a probe tube microphone assembly. Replicated measurements were made at five locations external to the pinna. They were: anterior-superior and posterior-superior positions simulating hearing aid microphone placement (locations 1 and 2) and 2, 4, and 6 cm lateral to the lateral edge of the pinna (locations 3, 4, and 5). Means, standard deviations, and ranges were compared, and statistical analyses were performed. The highest pSPL values were recorded lateral to the pinna, and the lowest values were obtained at the simulated hearing aid positions. ANOVAs indicated a main effect for pSPL, and post hoc testing demonstrated a significant difference between the posterior-superior and 2 cm lateral to the pinna positions. Variability was largest at the posterior-superior and 6 cm positions, and lowest 2 cm from the pinna. From this study, we concluded that pSPL and variability are both important criteria for selecting an optimal reference microphone site and both can affect the accuracy of ear canal measurements. A reference site 2 cm from the pinna eliminates attenuation of the signal and is the least variable site.