Self-monitoring of blood glucose levels has become an important instrument for the management of patients with diabetes mellitus. Both patients and physicians expect that the monitors will provide reliable results. Numerous environmental, physiologic, and operational factors can affect system performance, yielding results that are inaccurate or unpredictable.This study examined the effect of one factor--high altitude--on the performance of seven blood glucose monitoring systems. The following monitors were compared: two One Touch II; two One Touch Basic; two Reflolux II (Accu-Chec in the USA); two Glucometer 3; one Glucometer 2, and one Accutrend Alpha. Double blood glucose level values were compared with a controlled reference laboratory test value, which was unknown to the investigator until the end of the study because the study was double blind. Blood glucose values were obtained using each of the monitors in 200 patients; 150 with diabetes mellitus, and 50 healthy subjects.The One Touch monitors were the only monitors that reported adjusted straight lines (Y = a+bX) that were very similar for all three techniques. In addition, these adjusted straight lines are those closest to the ideal line, Y = X. These same monitors were the only ones that did not reject the null hypothesis Ho: a = 0. The relative deviation index at the 20% level was less than 3.5% for the One Touch II and One Touch Basic monitors; for the rest of the monitors, the index was over 14%. The clinically accepted EGA region was similar for all study monitors.In conclusion, the One Touch II and One Touch Basic Monitors showed greater accuracy in comparison to the other devices. The evaluation of the clinically acceptable region shows practical reliability for all of the monitors used.