Prior studies of oligonucleotide microarray-based mutational analysis have demonstrated excellent sensitivity and specificity except in circumstances where a frameshift mutation occurs in the context of a short repeated sequence. To further evaluate this circumstance, a series of nucleic acid samples having heterozygous mutations within repetitive BRCA1 sequence tracts was prepared and evaluated. These mutations included single nucleotide insertions and deletions in homopolymer runs, insertions and deletions of trinucleotide repeats, and duplications. Two-color comparative hybridization experiments were used wherein wild type reference and test targets are co-hybridized to microarrays designed to screen the entire BRCA1 coding sequence for all possible sequence changes. Mutations in simulated heterozygote samples were detected by observing relative losses of test target hybridization signal to select perfect match oligonucleotide probes. While heterozygous mutations could be readily distinguished above background noise in 9/19 cases, it was not possible to detect alterations in a poly dA/dT tract, small triplet repeat expansions, and a 10 bp direct repeat. Unexpectedly, samples containing (GAT)(3) triplet repeat expansions showed significantly higher affinity toward specific perfect match probes relative to their wild type counterparts. Therefore, markedly increased as well as decreased test sample hybridization to perfect match probes should be used to raise a suspicion of repetitive sequence changes.