Air leakages in food and ingredient packages which are sealed in vacuum environments may cause a marked deterioration of the product, leading to a loss of functionality. Manufacturers of such products have very stringent but rather costly quality control procedures and there is a pressing need for developing more economical ways of automated quality control techniques to test the vacuum packages reliably. However, due to the fact that the defect size of a typical package with a leakage problem could be micro- or nano scale, such faults are not detectable using conventional techniques. In this paper, the performance of a proposed acoustic method is assessed for the detection of air leakage in instant dry yeast packages sealed in a vacuum environment, which are typical of food and ingredients packaged under vacuum conditions. The investigation is carried out in both laboratory and in-situ environments. The acoustic pressure created by leaking air into the faulty packages is measured using a low-noise microphone in an acoustic chamber. Faulty packages are then identified using the changes in measured sound pressure levels within a certain frequency band. A mathematical model is also proposed to predict the pressure inside a yeast package with certain defect size as a function of time. The mathematical model is then used to determine the size of a defect causing the leakage, using the time required for the pressure inside a faulty yeast package to reach to a threshold level. The results of this investigation show that, using the state of the art measurement techniques, it is possible to detect packages with leakage problem if the diameter of the defect is greater than a few tens of micrometres. (C) 2018 Elsevier Ltd. All rights reserved.