In the recent times, the reduction of process variability has become very important for delivering cost-effective product. As per Taguchi's Loss Function, higher variability means higher cost also. The quantification of variability is a tremendous challenge to NVH engineers. To understand the NVH characteristics of an engine, measurements at various locations are needed. Moreover, each measurement is also a function of engine speed. In order to capture the process variability, multiple engines need to be measured. Thus, the final data is a function of engine, location, and speed. The objective of this paper is to define the various types of variability associated with the NVH data and to develop a procedure to estimate them. This paper contains the examples from the test data.