Background: Food allergy (FA) is an increasing health problem in children. Previous studies have reported conflicting results about the diagnostic value of atopy patch test in food allergies. The aim of the present study was to investigate the accuracy of atopy patch test in identifying non-IgE-mediated gastrointestinal FA to pasteurized/homogenized cow›s milk, heated raw cow›s milk, white and yolk egg, soy, wheat, walnut, sesame, pistachio, almond, peanut, date and cumin which are popular food ingredients in southern Iran. Methods: This study was performed on children with gastrointestinal allergic problems who were not improved after at least 4 weeks of cow’s milk protein elimination. Atopy patch test (APT) was performed and the elimination diet was considered according to positive skin results and after resolving symptoms, introduction of each accused food was done sequentially. Results: Fifty tree children under 7 years old with mean age of 19.6 months underwent APT and the results were verified by open oral food challenge (OFC). Sensitivity in the range of 59%-95%, specificity of 80.7-92.8%, positive predictive values of 75-96.4% and negative predictive value of 23%-80.7% were calculated depending on the type of food. Compared to the heated raw cow’s milk, the pasteurized/homogenized cow’s milk reaction was significant. Conclusion: APT can be included in the diagnostic workup of non-IgE-mediated GI allergy because it is safe and has great accuracy. However, several aspects require further investigation especially to enable the standardization of the technique. We should be aware of the allergenicity of our foods due to processing and geographic region.