Chinese hamster ovary (CHO) cell cultures used to produce biopharmaceuticals are tested for mycoplasma contamination as part of the ensurance of a safe and pure product. The current U.S. Food and Drug Administration (FDA) regulatory guideline recommends using two procedures: broth/agar cultures and DNA staining of indicator cell cultures. Although these culture methods are relatively sensitive to most species, theoretically capable of detecting as few as 1-10 cfu/ml of most species, the overall procedure is lengthy (28 d), costly and less sensitive to noncultivable species. The detection of mycoplasma using the polymerase chain reaction (PCR) method has been considered an alternative method because it is relatively fast (1-2 d), inexpensive, and independent of culture conditions, however, limitations in sensitivity (limit of detection >/=1000 cfu/ml) and the risk of false positive and false negative results have prevented PCR from replacing the traditional culture methods in the industrial setting. In this report, we describe a new PCR assay for mycoplasma detection that appears to resolve these issues while being sufficiently simple and inexpensive for routine use. This assay applies readily available techniques in DNA extraction together with a modified single-step PCR using a previously characterized primer pair that is homologous to a broad spectrum of mycoplasma species known to infect mammalian cell cultures. Analysis is made easy by the detection of only a single amplification product within a narrow size range, 438-470 bp. A high sensitivity and specificity for mycoplasma detection in CHO cell production cultures is made possible through the combination of three key techniques: 8-methoxypsoralen and UV light treatment to decontaminate PCR reagents of DNA; hot-start Taq DNA polymerase to reduce nonspecific priming events; and touchdown- (TD-) PCR to increase sensitivity while also reducing nonspecific priming events. In extracts of mycoplasma DNA, the limit of detection for eight different mycoplasma species is 10 genomic copies. In CHO cell production cultures containing gentamicin, the limit of detection for a model organism, gentamicin-resistant M. hyorhinis, is 1 cfu/ml. The sensitivity and specificity of this PCR assay for mycoplasma detection in CHO cell production cultures appear similar to the currently used culture methods and thus should be considered as an alternative method by the biopharmaceutical industry.