Carolina Power & Light Company has two nuclear power plants in operation, with a third soon to go into operation. Their nuclear power program is supported by a central health physics group that has as one of its responsibilities the upkeep of emergency air samplers for use at the plants when needed. A potential problem was identified in that the plants did not all use the same air sampling components such as particulate filters and radioiodine collection cartridges. The concern was that the substitution of components could change the air flow rates from the calibrated flow rates. This study was to investigate the calibration errors in the emergency air sampling program at CP&L's central health physics facility. Both high volume and low volume air samplers are maintained in operational status. The high volume samplers are used to sample airborne particulates and the low volume samplers are used for sampling radioiodine and particulates. The high volume samplers required little testing to determine that the total error in their flow rates was well within the recommended limit of +/- 20% specified in the REGULATORY GUIDE 8.25.(1) for sampled air volumes. In a worst case scenario the high volume samplers had a 10.5% total error. The low volume samplers required more testing than did the high volume samplers because of problems discovered during the initial testing. The major problem was air leakage in the sampling head on the sampler. It was found that this air leakage would be the worst during plant use of the sampler. Because of the leakage the calibrated air flow would not be the correct value. However by using the calibrated flow rate, the calculated concentration of airborne radionuclides would be higher than the actual concentration present. Recommendations were to use the high volume samplers with no changes. There were three possible choices given in the recommendations for the low volume samplers. The first is to replace the low volume samplers with new units that have desirable features that the present samplers do not have. The second choice is to replace only the sampling heads. The third choice is to suspend any further particulate sampling with the low volume samplers in their present condition if neither of the first two recommendations were implemented.