Sample stability is essential for reliable results in clinical laboratory practice. The aim of the work was to investigate the change in values of hematological indices in samples stored for up to 72 hours under different temperature regimes. A total of 60 whole blood samples stored under different conditions were analyzed: at room temperature (25°C), heated to 35°C and cooled to 4°C. Analysis was performed at different time points: immediately after blood sampling and then consecutively after 3, 6, 12, 24, 48 and 72 hours. K2EDTA anticoagulant tubes were used, and results were obtained using a UniCel DxH 800 hematology analyzer. The median shift of the parameters relative to baseline for each combination of time and temperature was assessed using the Wilcoxon matched pairs test. The shift in hemogram values obtained using Bland-Altman plots was compared with the maximum permissible error specified in the quality specification for the desirable error. Hemoglobin, erythrocyte count, mean erythrocyte hemoglobin content and platelet content were stable for at least 72 hours at all temperatures used in the experiment. For the other tested parameters, the first unacceptable changes in hemogram values were observed after 3 hours when the samples were stored at 25°C and 35°C. In samples cooled to 4°C, the first statistically significant differences were recorded after 6 hours. As a result, storage of samples for 72 hours at room temperature led to reliable unacceptable changes in 6 hemogram parameters of the 11 studied, at 4°C 5 parameters changed unacceptably, and at 35°C – 7 parameters. The obtained results, on the one hand, indicate that when analyzing the results of hematological tests is performed with a delay after sample collection, changes in hematological parameters should be considered; on the other hand, they provide information about the list of parameters subject to temperature-time changes, as well as about the intensity of these changes.