Undesirable levels of scrap and rework have been a chronic problem in the manufacturing environment. This study was conducted to determine the relationship of part dimension, and its process variability to manufacturing performance. Because manufacturing variability is usually evaluated in detection and attribute terms, a quantifiable and early prediction performance measure was desired. This study utilized a variety of statistical methods to establish a dimensional link to manufacturing performance so that a quantifiable indicator (dimension) could be evaluated and controlled for better manufacturing performance and yield. Additionally, the relationships understood from this study will provide significant design input for future product manufacturability of automotive instrument panels.