In order to evaluate the feasibility of using quantitative polymerase chain reaction (PCR) to evaluate gene dosage, we developed an assay to detect alpha-globin genes, which are frequently deleted in alpha-thalassemia patients. In this quantitative assay alpha-1 and alpha-2 globin consensus regions are coamplified by one oligonucleotide pair, along with a second primer pair targeting a single-copy reference gene, namely, tumor necrosis factor alpha, or TNF-alpha. A series of DNA samples titrating alpha-globin against TNF-alpha DNA have a strong linear relationship between template ratios and product ratios (r > 0.98). Minimal sequence divergence (91% homology) between alpha-1 and alpha-2, internal to the identical primer annealing sites, results in a lower amplification efficiency for alpha-1, to 94% of alpha-2 for each cycle. Furthermore, when applied to a variety of individual DNA samples, the signal ratios of alpha-globin to TNF-alpha were far more variable than previously observed for titrated control DNA. We conclude that DNA isolates from different individuals may have idiosyncratic changes in amplification efficiency owing to polymorphic sequence variation and/or variable presence of unidentified contaminants. Despite these potentially confounding factors, however, we were able to identify by quantitative PCR a single gene deletion later confirmed by Southern blot analysis in 20 individual DNA samples.