The etiology of mental retardation (MR), often presenting as developmental delay in childhood, is unknown in approximately one-half of cases. G-banding is the standard method for investigating those suspected of having a chromosomal etiology; however, detection of structural abnormalities is limited by the size and pattern of the G-bands involved. Rearrangements involving subtelomeric regions have been shown to cause MR and this has generated interest in investigating the prevalence of these rearrangements using telomere-specific probes. In addition, because cryptic interchromosomal rearrangements may not be small or confined to chromosomal ends, spectral karyotyping (SKY) using chromosome-specific painting probes may be of value. We report here a study using these two FISH-based techniques in 50 children with idiopathic MR or developmental delay and normal GTG-banded karyotypes. Our objective was to assess the prevalence of cryptic rearrangements in this population using subtelomeric FISH and SKY. Three rearrangements were detected by subtelomeric FISH: a derivative 5 from a maternal t(5;21); a recombinant 11 from a paternal pericentric inversion; and a 2q deletion that was also present in the mother. Only the derivative 5 was detected by SKY. SKY did not detect any interstitial interchromosomal rearrangement. The prevalence of clinically significant cryptic rearrangements by subtelomeric FISH and SKY was thus 4% (95% confidence interval 0.5-13.7) and 2% (95% CI 0.05-10.7), respectively. This study supports the view that G-banding does not detect all clinically significant chromosomal abnormalities and that subtelomeric FISH and SKY can detect some of these abnormalities.