Introduction: The Milan System for Reporting Salivary Gland Cytopathology (MSRSGC) was proposed by the American Society of Cytopathology and the International Academy of Cytology to bring uniformity in the reporting system and the treatment protocol. A wide range of risk of malignancy for each category has been reported by various authors by applying the system. Aim: We intend to study the cytohistological concordance and the ROM for each of the diagnostic categories of the Milan system. Materials and Methods: The study included 292 cases of fine-needle aspiration cytology (FNAC) of salivary gland lesions over a period of 3 years. The diagnosis of these cases was reclassified into the 6 categories of the Milan system. The cytohistological concordance and ROM for each category of the Milan system were calculated based on the clinical and histopathological follow-up. Results: The patients’ age ranged from 3 to 81 years with the mean of 42.65 ± 16.3 years. The cases included 189 (64.7%) parotid, 82 (28.1%) submandibular, and 21 (7.2%) cases of minor salivary gland swellings. Follow-up histopathological diagnosis for 102 cases was available. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated to be 64.28, 97.01, 90, 86.67, and 87.37%, respectively. After reclassification, the number of cases in each category was as follows: category I: 31 (10.62%), category II: 80 (27.4%), category III: 2 (0.68%), category IVA: 143 (48.97%), category IVB: 1 (0.34%), category V: 13 (4.45%), and category VI: 22 (7.53%). The calculated ROM was as follows: category I: 42.86%, category II: 26.67%, category III: 100% category IVA: 10.17%, category IVB: 0%, category V: 71.42%, category VI: 100%. Conclusion: FNAC is an excellent procedure to differentiate benign from malignant tumors, and MSRSGC is a useful system for risk assessment and deciding the further treatment protocol. Our findings also suggest that in addition to the surgical follow-up, inclusion of the clinical and radiological follow-up may be a better strategy for calculation of ROM, especially for categories I and II.