Introduction: The aim of this study was to compare ionized calcium (iCa) concentrations in arterial heparinized blood and venous serum and toinvestigate time-dependent variation of iCa in serum samples centrifuged and analysed at different times.Materials and methods: Ionized calcium was measured (N = 25) in arterial blood within 20 min after puncture, and in serum within 10 min aftercentrifugation conducted 30 min after sampling. Effect of time between sampling and centrifugation was examined in three tubes (N = 30) centrifuged15, 30 and 60 min after sampling, and analysed within 10 min. Effect of time between centrifugation and analysis was investigated in threetubes (N = 31) centrifuged 30 min after sampling and analysed: 0-10, 30-40 and 90-100 min after centrifugation. Ionized calcium was measured onthe Siemens RapidLab 348EX analysers. Statistical significance was tested using Wilcoxon test and ANOVA analysis. Clinical significance was judgedagainst reference change values (RCV).Results: No statistically significant difference was found between iCa in arterial blood and serum (P = 0.274). A statistically significant decrease wasfound: in tubes centrifuged 60 and 15 min after sampling versus 30 min (P = 0.005, P = 0.003); and in tubes analysed 30-40 and 90-100 min aftercentrifugation versus 0-10 min (P = 0.021, P = 0.027). Clinically significant changes were observed: 60 versus 30 min (centrifugation) and 90-100versus 0-10 and 30-40 min (analysis).Conclusions: Timely analysed arterial blood and serum samples can be used interchangeably. To avoid clinically significant variations, serum tubesshould be centrifuged within 30 min after sampling, and analysis should be performed within 30 min after centrifugation.