Approximately 95% of individuals with spinal muscular atrophy (SMA) lack both copies of the SMNt gene at 5q13. The presence of a nearly identical centromeric homolog of the SMNt gene, SMNc, necessitates a quantitative polymerase chain reaction approach to direct carrier testing. Adapting a radioactivity-based method described previously, multiplex polymerase chain reaction was performed using fluorescently labeled primers followed by analysis on an ABI 373a DNA sequencer. The SMNt copy number was calculated from ratios of peak areas using both internal and genomic standards. Samples from 60 presumed carriers (50 parents of affected individuals and 10 relatives implicated by linkage analysis) and 40 normal control individuals were tested. Normalized results (to the mean of five or more control samples harboring two copies of the SMNt gene) were consistently within the ranges of 0.4 to 0.6 for carriers (one copy) and 0.8 to 1.2 for normal controls (two copies), without overlap. Combining linkage analyses with direct carrier test results demonstrated de novo deletions associated with crossovers, unaffected individuals carrying two SMNt gene copies on one chromosome and zero SMNt gene copies on the other chromosome, and unaffected individuals with three copies of the SMNt gene. This report demonstrates that fluorescencebased carrier testing for SMA is accurate, reproducible, and useful for genetic risk assessment, and that carrier testing may need to be combined with linkage analysis in cer-tain circumstances. Am. J. Med. Genet. 85: 463-469, 1999.