Preexisting fossil collections, often the only source of data from sites that are no longer accessible, have been widely used in systematics but underutilized in quantitative paleontological studies. Here, the potential utility and limitations of materials obtained from preexisting collections is explored using a paleoecological case example (confamilial predation by drilling naticid predators). Specimens of the gastropod Cochlis pseudoepiglottina were collected from a single locality (Nidolères outcrop, Pyrénées-Orientales Département, southern France) of an unnamed, early Pliocene unit; the site is currently inaccessible. A statistical analysis of 183 specimens, representing three separate collections, indicates that these datasets are quantitatively consistent and provide interpretable numerical data on various aspects of drilling predation patterns. Despite various limitations, the results indicate that (1) intense confamilial predation affected C. pseudoepiglottina; (2) drilling attacks may have been size selective; (3) the frequency of failed attacks recorded by incomplete drill holes was high; and (4) the ratio of successful attacks declined dramatically with increase in snail size, suggesting a strong size-refuge effect. Our interpretation is hampered by the exclusion of the smaller size classes of C. pseudoepiglottina and a lack of bulk-sampled, assemblage-level quantitative data on all fossil taxa at this taxonomically diverse site. Nevertheless, the results provide a useful addition to previous literature on drilling predation, including quantitative data that are potentially applicable in meta-analytical compilations (e.g., prey effectiveness). Despite limitations, preexisting collections—especially those representing fossil sites that no longer exist or are inaccessible—can yield useful quantitative data.