A characteristic of Australian stable continental region (SCR) faults appears to be the temporal clustering of surface rupturing earthquakes. With the possible exception of active faults in the Flinders Ranges and Mt Lofty Ranges regions of South Australia, active periods of earthquake activity comprising a finite number of events are separated by much longer periods of seismic quiescence. This behaviour poses particular problems for seismic hazard assessment in that it implies that recurrence of large earthquake events is not random (Poisson) as is implicitly assumed in most seismic hazard assessment methods. We present a model that helps to conceptualise the points critical to understanding the hazard posed by intraplate faults and in modelling this hazard probabilistically. Specifically, we look to identify whether the SCR fault in question is about to enter an active period, is in the midst of an active period, or is in a quiescent period. In relation to this issue it becomes important to know how many previous ruptures the fault has generated in its current active period (should it be in one), and specifically, if a fault is in an active period, what is the average recurrence interval and what is the variability around this average. This average could be incorporated statistically into probabilistic seismic hazard assessments.