STUDY OBJECTIVES: By measuring the apnea length, ventilatory phase, respiratory cycle length, and loop gain, we can further characterize the central apneas of high altitude (CAHA). METHODS: Sixty-three drivers of all-terrain vehicles, working in a Peruvian mine located at 2,020 meters above sea level (MASL), were evaluated. A respiratory polygraph was performed in the first night they slept at high altitude. None of the subjects were exposed to oxygen during the test or acetazolamide in the preceding days of the test. RESULTS: Sixty-three respiratory polygraphs were performed, and 59 were considered for analysis. Forty-six (78%) were normal, 6 (10%) had OSA, and 7 (12%) had CAHA. Key data from subjects include: residing altitude: 341 +/- 828 MASL, Lake Louise scoring: 0.4 +/- 0.8, Epworth score: 3.4 +/- 2.7, apneahypopnea index: 35.7 +/- 19.3, CA index: 13.4 +/- 14.2, CA length: 14.4 +/- 3.6 sec, ventilatory length: 13.5 +/- 2.9 sec, cycle length: 26.5 +/- 4.0 sec, ventilatory length/CA length ratio 0.9 +/- 0.3 and circulatory delay 13.3 +/- 2.9 sec. Duty ratio media [ventilatory duration/cycle duration] was 0.522 +/- 0 0.128 [0.308-0.700] and loop gain was calculated from the duty ratio utilizing this formula: LG = 2pi / [(2piDR-sin(2piDR)]. All subjects have a high loop gain media 2.415 +/- 1.761 [1.175-6.260]. Multiple correlations were established with loop gain values, but the only significant correlation detected was between central apnea index and loop gain. CONCLUSIONS: Twelve percent of the studied population had CAHA. Measurements of respiratory cycle in workers with CAHA are more similar to idiopathic central apneas rather than Hunter-Cheyne-Stokes respiration. Also, there was a high degree of correlation between severity of central apnea and the degree of loop gain. The abnormal breathing patterns in those subjects could affect the sleep quality and potentially increase the risk for work accidents.