Plasma-derived (pd) and recombinant (r) factor IX (FIX) differ in pharmacokinetic (PK) properties. These differences and their clinical implications have been debated since the introduction of rFIX. The aim of this review was to describe the comparative disposition of pdFIX and rFIX and will for this purpose begin with an overview of population PK modelling. In contrast to the model-independent method, a population PK model can analyse sparse data sets obtained in various settings, provide parameter values that can be used to predict coagulation factor levels with any kind of single or multiple dosing and include statistical analysis of variation between individuals. Population modelling has also clearly demonstrated the difference in PK between pdFIX and rFIX. Their distribution characteristics influence the FIX coagulant activity (FIX:C) level vs. time curve during the early hours after infusion. In vivo recovery and elimination half-life are consequently not adequate descriptors of the effective PK of FIX, and for new analogues with modified PK, differences in distribution might be clinically important. Calculated doses to maintain 1% trough levels during twice-weekly prophylactic treatment are considerably higher with rFIX than with pdFIX and roughly correspond to dosing in clinical studies. However, the putative relationship between FIX:C trough level and therapeutic outcome has never been confirmed in a clinical trial. Comparative studies on prophylaxis with different types of FIX are needed.