Acinetobacter baumannii is a clinically significant pathogen found in both environment and clinical samples. Its accurate identification is crucial for infection control and patient management. This study aims to assess the effectiveness of the commercially available Leeds Acinetobacter Medium (LAM) in the presumptive identification of A. baumannii, and compare its performance with that of multiplex PCR (mPCR) for accurate identification. A. baumannii was isolated using LAM from water (n = 100) and clinical respiratory samples (n = 100) which included endotracheal tube samples, sputum samples, etc. and the accuracy of identification was compared with mPCR. Additionally, the cultural characters of known A. baumannii strains (n = 42) were evaluated on LAM and the isolates were tested with mPCR. A. baumannii prevalence was 5.5% in water and 35.1% in clinical respiratory samples. LAM’s identification accuracy, compared to mPCR, showed 66.7% sensitivity and 76.5% specificity for water samples, and 88.5% sensitivity and 37.5% specificity for clinical respiratory samples. Positive predictive values (PPV) were 14.3% and 43.4%, while negative predictive values (NPV) were 97.5% and 85.7% for water and clinical respiratory samples, respectively. Sequential testing involving preliminary identification followed by LAM culture achieved perfect agreement (kappa = 1) and higher sensitivity/specificity (100%) between LAM and mPCR. The study established mPCR as a reliable standard for A. baumannii identification, while preliminary identification tests such as Gram’s stain, catalase, oxidase and motility test improved LAM’s accuracy by reducing false positives. Limitations in A. baumannii identification were observed with LAM, suggesting its use for screening samples rather than for definitive identification.