The findings show that the data-driven approach adopted in this research is robust; the magnitude of likelihood ratios (statistical measure that offers a balanced presentation of the strength of the evidence) obtained under both propositions used focuses on the major evidential contribution that small portions of fingermarks (three minutiae) can provide, regardless of their position on the general pattern. The Tippett plots computed for all regions together and separately for each of the zones show slight differences between them. The comparison between the two datasets through the rates of misleading evidence does not suggest a problem in relation to a change in database. The authors envision expanding the proposed model to deal with more than three minutiae through a study of a larger collection of friction-ridge skin impressions. The within-source variability was estimated by using a sample of 216 fingerprints from 4 fingers (2 donors). Between-sources variability involved a database of 818 ulnar loops from randomly selected males. Features vectors used for statistical analysis were obtained through a preprocessing step based on Gabor filtering and image processing that extracted minutia position, type, and direction. Spatial relationships were coded by using Delaunay triangulation. The metric used to assess similarity between two feature vectors was based on a Euclidean distance measure. 3 tables, 16 figures, and 31 references