Ultrasonic imaging technologies are rapidly being transitioned to the production environment. An example of this is occurring in the aerospace industry, where digital data acquisition and imaging are being used to improve the ultrasonic inspection of large grained alloys. [1] The availability of digital data and ever increasing computing power opens the door for more sophisticated data analysis techniques than have been used in the past. Such potential techniques include the Wiener filter to improve resolution, dynamic thresholding to improve detection, signal-to-noise (SNR) based material acceptance criteria, and the estimation of the probability of detection (POD) of a given inspection. [2–5] An element critical to the success of all these techniques is an accurate estimate the distribution of the ultrasonic reflections from grain boundaries which are commonly referred to as grain noise. This paper presents a technique to estimate the parameters of closed-form statistical distributions from grain noise data and analyzes the quality of the fit of several distributions to the grain noise found in ultrasonic images of titanium alloys.