Two episodes of acute illness occurring within a short time in a factory manufacturing rotary compressors were described. A malfunction in a regulator valve caused an explosion in the factory on January 29, 1986, with no injuries reported at the time. When the building was reopened on February 3 a degreaser malfunctioned. This resulted in workers in the assembly room being exposed to trichloroethylene (79016) concentrations of 224 parts per million (ppm), well above the OSHA standard of 100ppm. Ten employees became acutely ill reporting symptoms such as headache, dizziness, disorientation, eye and throat irritation, vomiting, dyspnea, and chest pain. Blood carboxyhemoglobin (COHb) concentrations determined in five workers were above normal, ranging from 13.3 to 14.8 percent. The workers were treated and released and the factory was closed for the day. Between February 4 and 18, 15 employees were seen at the shop infirmary complaining of headache, eye and throat irritation, nausea, and muscle aches. Blood samples analyzed by local medical facilities had COHb concentrations of 13.7 to 20.0 percent. These were later determined by NIOSH to be inaccurate. Air sampling found carbon-monoxide (630080) exposures to be well below the NIOSH recommended limit of 35ppm. Another outbreak of illness involving 19 workers occurred on February 19. Seventeen were tested for blood COHb and all were within the normal range 0.0 to 8.0 percent. Screening for trichloroethylene, fluorocarbons, and methylene-chloride was also negative. Carbon-monoxide concentrations in the factory that day were below 15ppm. The factory was reopened after an open meeting was held on February 23 in which NIOSH informed the employees of their finding and assured that the factory environment was safe. A questionnaire survey on February 25 found that during the first outbreak the ill workers were 2.26 times as likely to enter the assembly room as the asymptomatic workers; however, during the second outbreak the cases did not enter the assembly room any more frequently than the healthy workers. No further outbreaks have occurred since the open meeting. The authors conclude that the second outbreak was a collective stress reaction resulting from the explosion, earlier toxic exposures, and the misleading blood COHb results.