Purpose: To validate two Continuous Glucose Monitoring Systems (CGMS) against arterial AccuChek® point-of-care glucose measurements in critically ill patients. Methods: Two CGMS, Guardian RT® (GRT) and FreeStyle Navigator® (FSN), were used for 5 days in mechanically ventilated patients with Multiple Organ Dysfunction Syndrome. The measurements at T=0, T +10 min and T + 20 min were linked to T=0 reference data to investigate a possible sensor delay. Data were analyzed by: 1) the absolute difference, 2) correlation, 3) Bland-Altman analysis, 4) Clarke Error Grid analysis. The practical features and effect of vaso-active medication on CGMS accuracy were evaluated. Results: In 350 paired measurements in 18 patients, the Median Absolute Difference was 10.6% to 11.6% for FSN® and 10.3% to 11.5% for GRT®. Pearson's correlation coefficients were 0.79 to 0.82 for FSN® and 0.76 to 0.80 for GRT®. Limits of agreement for Bland Altman analysis varied between -2.49 to 1.99 for FSN® and between -2.59 to 2.67 for GRT®. Of the FSN® measurements, 96.9% - 97.2% were in zones A+B of the Clarke Error Grid, versus 97.6% - 97.8% for GRT®. Both CGMS performed slightly better at T +20 min compared to T=0 measurements. Seven hypoglycaemic events were detected by both CGMS systems and not by routine AccuChek® measurement. The GRT® system had fewer practical problems compared to the FSN® system. No influence of vasoactive drugs was found. Conclusions: Both systems performed equally well. Using a CGMS can show early detection of hypoglycaemia. A disadvantage of the FSN® is the 10-hour warm-up period it requires. © 2010, Nederlandse Vereniging voor Intensive Care.