Our understanding of FMR1 trinucleotide instability has increased dram
atically with knowledge of its detailed structures. While most arrays
seem to be protected by interspersions, for a few the price of perfect
ion is instability. Although there remain many unanswered questions, d
iagnosis in the ''grey zone'' can be greatly improved by studying arra
y content. For the future, as we strive to delineate normal from premu
tation, we should increasingly be able to estimate rates of instabilit
y for future generations and predict the risk of conversion to the ful
l mutation.