We have evaluated serum cholesterol in a primary care center with a recently introduced portable analyzer: Reflotron (Boehringer Mannheim). The linearity and exactitude were investigated with Hitachi 704 (Boehringer Mannheim) used in the reference laboratory. Accuracy was also tested. We found a good accuracy after repeated analysis (n = 30) of two control sera with a variation coefficient of 2.77% in the first (mean = 125.14 mg/dl, SD 3.47) and 2.94% (mean = 254.53 mg/dl, SD 7.49) in the second. We found an overall lack of accuracy of Reflotron with a more pronounced negative bias in capillary blood (mean d = 30.93 mg/dl, dSE 3.25) than in serum (mean d = dSE 1.12). We found a positive linear correlation both using serum (r = 0.9845; n = 45) and capillary blood (r = 0.9094; n = 30). We feel that Reflotron is useful for primary care due to its good accuracy and linearity regarding the reference method, but its inexactitud should be taken into account. We emphasize the need for quality control to investigate its limitations when the results are analyzed.