The efficacy of positive pressure respirators under field conditions during actual fire fighting activities was evaluated at five major fire departments in the United States. Probes placed into the respirators continuously sampled for contaminants both inside of and outside of the facepiece of the respirator. Fourier Transform Infrared Spectrophotometry was used to subsequently identify the contaminants. Individual measurements were made on 64 firefighters at 25 active fire scenes. Preexposure and post exposure carbon- monoxide (630080) breath samples were collected. Only low levels of contaminants were found in each of the fires sampled, making measurement of protection impractical. An additional problem was the inability to determine the test subjects' prior exposure to carbon-monoxide. The types of fire studied in this research included rubbish outside, building and contents occupied, unoccupied building, passenger vehicle, unoccupied building fought from outside, building and contents unoccupied, and oil from a generator. Removal of the respirator was a significant limiting factor in respirator effectiveness.