How well do the quality assurance tests for latex condoms detect the potential for virus penetration? Defective condoms were created by puncturing with small acupuncture needles (120, 160 or 300 μm diameter) and were then tested by three quality assurance (QA) tests (the water leak test, a proposed electrical test, and the air burst test) and by a virus penetration test. Punctures (short tears) were detected by the FDA water leak test at the rate of 52% for punctures in the condom body and 9% in the condom tip; by the proposed ISO electrical leakage test (28% and 18% for body and tip, respectively); or by the ISO air burst test (30% and 5%, respectively). A standardized in vitro test of virus penetration under physiologic-based detected 100% and 48% of the punctures for body and tip, respectively. Thus the small tears created by puncture may not be detected in quality assurance tests, but would allow some virus penetration in the laboratory virus penetration test. A puncture in the tip of a condom was most likely to escape detection by each of the test methods,presumably because a puncture in that location does not open up as easily as elsewhere. Data with the air burst test indicated that raising the burst volume and/or burst pressure pass/fail limit would increase the likelihood of detecting a puncture defect to >93% when in the condom body, but to <15% when in the tip.