Non-random tumour-specific chromosomal abnormalities have been observed in cells of many different human tumours. In Wilms' tumour (WT) and retinoblastoma, a chromosomal deletion occurs germinally or somatically and has been considered an important step in tumour development. One class of potential cellular transforming genes comprises the cellular homologues of the transforming genes of highly oncogenic retroviruses. A remarkable concordance between the chromosomal location of human cellular oncogenes and the breakpoints involved in acquired chromosomal translocations is becoming apparent in various cancers: the oncogenes c-mos, c-myc and c-abl are located at the breakpoints that occur in acute myeloblastic leukaemia, Burkitt's lymphoma and chronic myelocytic leukaemia respectively. Thus when the oncogene c-Ha-ras1 was localized to the short arm of human chromosome 11 (refs 6-8; region 11p11 leads to p15 and not 11p13 as stated in ref. 5), it was proposed as a possible aetiological agent in the aniridia-WT association (AWTA) that results from a deletion of 11p13 (although a transforming gene recently isolated from a WT cell line (G401) was shown not to be homologous to either c-Ha-ras or c-Ki-ras9). We have now looked for deletion or rearrangement of c-Ha-ras1 in the DNA from four subjects with del(11p13)-associated predisposition to Wilms' tumour, aniridia, genitourinary abnormalities and mental retardation. We report here that in no case is c-Ha-ras1 deleted, and we have further refined its location to 11p15.1 leads to 11p15.5. On the basis of enzyme studies and direct gene dosage determination for c-Ha-ras1 and beta-globin in neoplastic and non-neoplastic tissues from one patient, we conclude that deletion of the normal counterpart of 11p cannot account for the development of the tumour.