Eighty seven referrals with Prader-Willi syndrome and 49 with Angelman's syndrome were studied. High resolution cytogenetics was performed on all probands. Molecular studies, performed on the proband and both parents in each case, utilised multiple probes from within and distal to the 15(q11-13) region in order to establish the presence of DNA deletion or uniparental disomy. In addition, FISH, with probes at D15S11 and GABR beta 3 from the Prader-Willi syndrome/Angelman's syndrome region, was performed on a subset of 25 of these patients. In the referral group with Prader-Willi syndrome, 62 patients had a normal karyotype and 25 were deleted on high resolution cytogenetics. Twenty nine were found to be deleted with DNA techniques. In the Angelman's syndrome group, 37 had a normal karyotype and 12 were deleted on high resolution cytogenetics while 26 were deleted on molecular studies. The diagnosis was reassessed in 35 referrals with Prader-Willi syndrome and 11 with Angelman's syndrome following a non-deleted, non-disomic result. Of individuals who were neither deleted nor disomic on DNA studies, a false positive rate of 11.4% (4/35) for Prader-Willi syndrome and 16.7% (2/12) for Angelman's syndrome was found for a cytogenetically detected deletion. The false negative rate for deletion detected on high resolution cytogenetics was 19.5% (12/62) for Prader-Willi syndrome and 35% (13/37) for Angelman's syndrome. Thus high resolution cytogenetics was shown to be unreliable for deletion detection and should not be used alone to diagnose either syndrome. There were no discrepancies with FISH in 25 cases when FISH was compared with the DNA results, indicating that FISH can be used reliably for deletion detection in both syndromes.