This research evaluated an offline vacuum decay leak detection system for 1,775-ml polyethylene terephthalate (PET) bottles. These bottles were filled with water and pulped and unpulped orange juice and induction sealed with an aluminum liner and an outer 38-mm continuous thread polypropylene cap. The objectives of this study were to evaluate (i) minimum leak size sensitivity of the instrument; (ii) ability to identify weak but nonleaking seals; (iii) effect of varying fill heights on the equipment's sensitivity; and (iv) percentage of false-positive and negative results likely to be obtained during a normal test run. To meet these objectives, leaks 5, 10, 15, 20, 30, 40, and 50μm were created in the PET bottles. A second set of bottles was induction sealed at high voltage and 1, 1.5, 2, 2.5, 3, 3.5, and4sof dwell time. A third set of bottles with good seals was filled with differing headspace measurements of brimful, 1, 2, 3, 4, and 5 cm. After optimizing the equipment, leak tests on random sets of leaking and nonleaking bottles showed 0.0% false-positive and 0.0% negative identifications. Results showed 5-μm minimum leak size detection for bottles filled with all products. Optimum seal conditions were >2 but <3s at high voltage. Product fill heights >2to ≤3 cm did not affect the efficiency of the equipment. These results show that this vacuum decay system has potential for use in identifying leaks in PET bottles used for food packaging.