BACKGROUND In this study the analytical performance of eight glucose point-of-care testing (POCT) devices was evaluated. For this purpose, POCT measurement of glucose in heparinized blood collected from patients was paralleled by determination of the glucose concentration in the respective plasma by an analyzer (Hitachi 917) in the central laboratory, providing traceable results. METHODS Trueness of POCT measurements was studied by comparing the plasma POCT values (mean of five measurements) with the results from the traceable measurement procedure (TMP). RESULTS The percentage of POCT results within +/-6% of the TMP mean value ranged from 24% to 50%, depending on the POCT device. Within the reference interval of plasma glucose (4.4-6.0 mmol/L), up to 67% of the POCT values were lower than 4.4 mmol/L, leading to a false diagnosis of hypoglycemia. In the hypoglycemic range (<4.4 mmol/L) up to 29% of the POCT analyses were false normoglycemic. CONCLUSIONS In conclusion, this study shows an insufficient trueness of glucose measurements by POCT devices in the normo- and hypoglycemic range. To improve quality assessment, sample splitting and simultaneous measurement of blood glucose concentration every 4 weeks by POCT devices and of plasma glucose concentration by a reliable TMP is recommended.