The earthquake catalog in southern Iceland between 20.25° W and 21.3° W is complete down to approximately Mw 0 and shows no evidence for magnitude shifts as a function of time since 1991. This means that 6902 earthquakes (M ≥ −0.1, depth ≤ 20 km) are available for studies of seismicity patterns preceding the two MS 6.6 mainshocks in 2000, in their source volumes and vicinity (20.25° W and 20.9° W). The power law of the frequency-magnitude relationship holds down to M0. Detailed mapping (including 3D mapping) of the b-value of the frequency- magnitude distribution shows that variations from b = 0.6 to 1.4 exist over distances of about 2 km. This suggests that the tectonic fabric in southern Iceland is heterogeneous on a small scale. The hypocenters of the two MS 6.6 mainshocks and one additional M 4.5 mainshock are associated with low b-values (0.6–0.8), statistically significantly different from volumes in their vicinity. Therefore, one can expect future mainshocks in southern Iceland to also emanate from volumes characterized by low b-values. However, the small dimensions of these asperities may render it difficult to identify them before the mainshocks occur. Low values of estimated local recurrence times (500–2000 years) map the two fault zones that ruptured in 2000 against a background of local recurrence times longer than 5000 years. The first of the two M 6.6 mainshocks emanated from a volume of short estimated local recurrence time. This supports the hypothesis that asperities with short local recurrence times control locations of major ruptures. Mapping of b-values in cross sections shows anomalies of high b at the bottom of the seismogenic crust, correlating with the change of its thickness in the middle of the study area. These high b-value anomalies, and other pockets of such anomalies at shallower depths, are best interpreted as caused by high pore pressure because evidence of fluids under high pressure is ubiquitous in this area.