Background Serum samples should be centrifuged for at least 10 min at 1300-2500 × g. Changed centrifugation conditions could compromise sample quality. The objective of this study was to compare the serum quality and turnaround time (TAT) using different centrifugation conditions. Methods The study was done in four different periods (A, B, C and D) at different conditions: for 10, 5 and 7 (A, B and C, respectively) at 2876 × g, and 7 (D) min at 4141 × g. Sample quality was assessed as the proportion of samples with: (a) aspiration errors, (b) H index >0.5 g/L and (c) suppressed reports of potassium (K) due to hemolysis. TAT was calculated for emergency samples. The proportions of samples (a), (b) and (c) were compared according to period A. Results The number of aspiration errors was significantly higher in samples centrifuged at 2876 × g for 5 min (p = 0.021) and remained unchanged when centrifuged for 7 min (p = 0.066 and 0.177, for periods C and D, respectively). In periods B, C and D, the proportion of samples with hemolysis was higher than that in period A (p-values 0.039, 0.009 and 0.042, respectively). TAT differed between all periods (p < 0.001), with the lowest TAT observed for B and D. The lowest number of samples exceeding 60-min TAT was observed in period D (p = 0.011). Conclusions The integrity of serum samples is changed with different centrifugation conditions than those recommended. Our study showed that shorter centrifugation at higher force (7 min at 4141 × g) significantly decreases TAT, with unchanged proportion of samples with aspiration errors.