Sample centrifugation is an essential step in the coagulation laboratory, as clotting tests are typically performed on citrated platelet (PLT) poor plasma (PPP). Nevertheless, no clear indication has been provided as to whether centrifugation of specimens should be performed with the centrifuge brake set to on or off. Fifty consecutive sodium citrate anticoagulated samples were collected and divided into two aliquots. The former was centrifuged as for Clinical Laboratory Standards Institute (CLSI) guidelines with the centrifuge brake set to on, whereas the latter was centrifuged again as for CLSI guidelines, but with the brake set to off. In the PPP of all samples, a PLT count was performed, followed by the analysis of activated partial thromboplastin time (APTT), prothrombin time (PT) and fibrinogen (FBG). The PLT count after samples centrifugation was substantially reduced, either with centrifuge brake set to on or off (5 ± 1 versus 3 ± 1 × 10/l; P = 0.009). The frequency of samples exceeding a PLT count less than 10 × 10/l was nearly double in samples centrifuged with the brake on than in those with the brake off (14 versus 8%; P < 0.01). Although no significant difference was found for APTT values, PT was slightly prolonged using the centrifuge brake set to on (mean bias 0.2 s; P < 0.001). FBG values were also significantly higher using the centrifuge brake set to on (mean bias 0.29 g/l; P < 0.001). The results of this study indicate that sample centrifugation for routine coagulation testing should be preferably performed with the centrifuge brake set to off for providing a better quality specimen.