Background: In Ethiopia, modern medicine was started in 1886 G.C and health professionals have been treating patients using different diagnostics algorithms, but there is a limited practice of examining post-mortem autopsy results and thinking back on the certainty of the diagnosis made for the patients before death. Conducting diagnostic discrepancy studies is important to understand the gap that exists in the clinical diagnosis and modify the diagnostic algorithms to give better and appropriate treatment. Methods: In this research, a formal, non-consecutive clinical case series study of under-five deaths included in the autopsy study between October 2019 and April 2022 was conducted. The premortem clinical diagnosis and post-mortem pathological diagnosis of the lung was compared for each case, and 02 physicians evaluated the results separately for concordance and discordance of diagnosis. McNemar test was used to see if there is a statistically significant difference between the premortem clinical diagnosis and post-mortem pathologic diagnosis of each case. P-value < 0.05 was considered statistically significant. Cohen’s kappa was also calculated to see the inter-rater variability. The reason for concordance and discordance in diagnosis was analyzed case by case for all under-five deaths included in this research by 02 separate senior physicians. Results: The magnitude of diagnostic discrepancy was 34% (95% CI: 20%-47%) and Cohen’s kappa Results was: 0.079. Reasons contributing to diagnostics discrepancy were: Case analysis problem (knowledge gap or compression problem) in 35.5%, Problem in consultation and teamwork in 35.5%, Documentation problem in 10%, Unavailability of diagnostic investigations in 8%, Communication gap between health facilities in 6%, Access to health facility in 3% and Work culture in 1%. Conclusions: The magnitude of misdiagnosis of cases is very high, and interventions should be implemented on the most common causes of diagnostics discrepancy to yield a better diagnosis of diseases by clinicians.