This study was designed to evaluate and compare the ability of respirator wearers to detect qualitative respirator fit test agents (saccharin and Bitrex) when the respirators were modified to include fixed size leaks. In recent years the number of persons who require fit testing has increased, partly in response to the needs of health care workers with potential exposure to infectious bio-aerosols. Many health care providers have chosen qualitative respirator fit testing using saccharin and/or Bitrex for a variety of reasons, including (but not limited to) low initial equipment cost. Respirators were modified to include a mid-line sampling probe between the nose and mouth for quantitative fit testing with a TSI PortaCount. A second modification included the introduction of a shortened 14-gauge intravenous catheter at the bridge of the nose. The fixed leak was designed to produce fit factors < 100 when unplugged, with an average fit factor of 67 among 26 respirator wearers. A complete fit test was not performed, because one purpose of this study was to determine the ability of respirator wearers to detect a known fixed leak during a single normal breathing exercise, without introducing unknown and potentially variable size leaks. Sensitivity threshold screening included a placebo and requirement to correctly characterize the taste of the agent used. Quantitative fit factors without leaks ranged from 96 to > 20,000 and 22 to 160 with the leak present. Twenty four of 26 subjects had fit factors < 100 (92%) when fixed leaks were induced. All subjects correctly detected Bitrex with fixed leaks (sensitivity = 100%). Nine of 26 subjects (35%) were unable to detect saccharin in the presence of a known fixed leak even though the average fit factor for these subjects was 77. When the two subjects with fit factors > 100 were excluded, only 16 of 24 respirator wearers were able to detect saccharin with fixed leaks (sensitivity = 67%). There were several important aspects of our study design worth noting, including the introduction of a placebo during sensitivity threshold testing, limiting the subject response time to a single maneuver, using a higher concentration of Bitrex than commercially available, and requiring the subjects to correctly characterize the taste of the qualitative test agent. In conclusion, leak detection was correctly identified with Bitrex, but not saccharin.