Two groups of white normal subjects (each containing 50 males and 50 females) were compared in respect to the frequency of the whorl pattern in their fingerprints. Each group was selected on the basis of having a name of British (surnames common in the British Isles) or Italian origin. In every other respect the selection was done at random from a large file of fingerprints of approximately 20,000 county employees from the same geographic area. The frequency of whorls in these ‘ethnic’ groups was found to be as follows ‘British’: males 22.4%, females 20.8%; ‘Italian’: males 34.4%, females 36.2%. This difference between the two groups is statistically significant and larger than that reported between patients and normal subjects in some studies. These results suggest that a population in a given geographic region in the USA may still contain distinct ethnic subgroups whose relative size may be different in patient and normal samples. Therefore, the practice of defining subjects only by geographic area and skin color is unsatisfactory and more attention should be paid to ethnic origin when matching patients and controls.