Evaluation of the leakage rate is necessary for infrastructure where hermetic performance holds critical importance. This study conducted a test to evaluate the air leakage rate of cracked reinforced concrete panels. Hollow beam specimens with an air chamber in the web were prepared to provide a relatively simple setup for airtight pressure testing under 200 kPa pressure. While testing, the lower flange of the beam specimen was subjected to tension cracking under transverse loading, and air pressure was applied to the air chamber of the beam to measure the air leakage rate through the flange plate as the magnitude of loading increased. The test variables were the differential air pressure, flange plate thickness, and reinforcement ratio. The test results showed that the leakage rate increased in proportion to the differential pressure. When the reinforcement ratio increased from 0.7 to 1.3 %, the leakage rate decreased by 90 %. When the plate thickness increased from 75 to 100 mm, the leakage rate decreased by 91 %. The analysis of the test results showed that the leakage rate was proportional to the cube of the crack width (R2 = 0.91–0.95), which agreed with existing prediction models. When the crack width at rebar location was used, the existing models showed the best predictions for test results. This result indicates that to accurately predict the air leakage rate of reinforced concrete members, the critical crack width at the location of reinforcement (i.e. the smallest crack width in the member thickness) is more important than the crack width at the member surface.