The Institute for Legal Medicine in Hamburg carried out a retrospective analysis for identity checks of stored blood samples. A total of 70,597 blood alcohol examinations were carried out between 1984 and 1993. Identity checks were commissioned for 187 people (0.26% of the total of all blood samples). Amongst those, 33 cases of "non-identity" were established (17.6% of all checked blood samples). The results are compared to an earlier study by Kleiber over a span of 10 years--between 1974 and 1983. While the frequency of identity checks decreased, the cases of "non-identity" increased quite drastically in comparison. In the cases where "non-identity" could be detected, it was not a matter of blood samples getting mixed up. However, it was a matter of intentional, sometimes very cunning ploys by means of other people's IDs, forged IDs and impersonations. In these cases the later checking of the blood group systems proved to be important evidence in the criminal proceedings. This evidence would not be available if there were to be no blood alcohol tests.