The use of statistics and probability in assessing DNA evidence is well known. The incorporation of statistical design, methodology, and uncertainty in other forms of forensic evidence is much less frequent, especially pattern evidence (fingerprints, ballistics matching, hair comparisons) which often arises in forensic evidence (and will be covered in depth by later speakers). In this talk, I will discuss the role of statistics in (a) developing quantitative procedures for pattern evidence based on science from other disciplines (e.g., imaging); (b) approaches to identifying sources of variability in such procedures; (c) experimental designs for comparing accuracy rates from "previous" versus "proposed" procedures; and (d) the role of traditional concepts from statistical process control in laboratory procedures.