Significant advances in learning outcomes are being demanded of all forensic disciplines. This is particularly true of the forensic identification sciences, including the analysis and assessment of footwear impression evidence. In 2009, the National Academy of Sciences reported a deficit in knowledge concerning the evidentiary value of forensic shoeprint impressions (NAS 2009). The conclusions of this review body support continued studies to extend knowledge concerning the statistical assessment of similarity, and the relative frequency of class and accidental characteristics present in various populations. To address these mandates, the proposed project asserts that an appropriately constructed research study can have a 4-fold impact on the field of shoeprint impression evidence.To achieve this impact in a reasonable time frame, semi-automated procedures and metrics are sought, allowing for the rapid analysis of hundreds of exemplar and questioned prints. In support of this effort, the proposed research will design, implement and validate customized code required to rapidly extract relevant similarity and frequency estimates from shoeprint data. The end result will be a prototype program that can be easily utilized to (1.) determine a quantitative measure of similarity between patterns based on correlation or other appropriately determined comparison methodologies, (2.) facilitate training and proficiency efforts (e.g. databasing of exemplar and test impressions), (3.) help articulate findings and conclusions, and (4.) rapidly extract the relative probability of the chance occurrence of accidentals that share similar characteristics or positional information. This latter benefit should support, at minimum, long-term investigations regarding fundamental phenomenology such as the spatial prevalence/dominance, development, erosion, and coincidental association of accidental features and patterns in random and specific populations To achieve these goals, more than 400 exemplar and questioned prints will be collected and compared over a 2.5 year time period. An accidental map of each shoe will be generated and used to populate descriptive feature vectors. Comparison of these vectors will result in similarity scores for known match and known non-match comparisons, providing a statistical assessment of the discrimination potential of accidental features and patterns that vary in both quality and totality. The results will deliver population frequency estimates regarding class and individual characteristics, the potential for false positives and negatives as a function of the number and type of accidentals present on an outsole, as well as recommendations regarding the use of automated and quantitative metrics of comparison for shoeprint evidence (e.g. current capabilities and future research initiatives such as variation in examiners' conclusions). In addition to the above, support of the proposed research will generate a long-term investment in the academic careers of three graduate students, and the opportunity for significant experiential learning at the undergraduate level.