For gaseous fire extinguishing systems, the maximum percent of agent in pipe, i.e., pipe volume vs. agent liquid volume should be determined for proper system design and performance by confirming the maximum length of pipe run in which their flow calculation methods can predict the discharge pressures and agent concentration. It is the purpose of this paper to determine the ability and limitations of the NFPA 12 flow calculation methodology to identify the maximum percent of agent in pipe by conducting full scale low-pressure CO2 system discharge tests. A total of twenty low-pressure CO2 system discharge tests were conducted under different conditions. If all the measured pressures at the three node points of pipe runs and the measured CO2 concentrations in the test enclosures did not deviate from the predicted values of computerized flow calculations by more than ¡¾10 percent, the tests were judged to be acceptable. In the test results, the low-pressure CO2 system with a pipe run exceeding 492 ft (150 m) was not likely to achieve the concentration required for fire extinguishment within the determined discharge time although the pipe network was installed in compliance with the calculations based on the pressure drop equation in NFPA 12.