[Comments on the international classification of oncologic diseases].

The original English version of ICD-O was slightly revised in German edition. Nevertheless, following inaccuracies remained: 1. Not other specification (NOS) should point but to the fact that there exist further possibilities for closer determination of coded tumour. That is why the abbreviation should not be used where other specification is concerned e. g. a histologenetical one. If the abbreviation is used instead of a code covering special forms (SF) they are to be specified by name. 2. Tumours with uncertain biological behavior are not coded consistently. They should be signed by 1 in denominator, not by word e. g. adding NOS. An idea is not allowed that increasing malignancy results in closer determination of tumours. 3. Isomorphic groups of tumours are not confined sufficiently by sequence of codes and a reserve is recommended to presumed absolute relation of code and unique localisation. If there are doubts it would be better keep code open as for localisation. 4. Real errors are in confusing synonyms especially of papillomas, in choice of the first-listed term, code duplicity (tumour: non-tumour), separation of diffuse and nodular tumours, and in incorrect translation. Most insufficiences could be avoided by proper comments which cannot solve obvinous conceptual and real errors.