Changes in the frequency of moderate-size events before and after the 28 June 1992 Landers earthquake are investigated, and their implications are discussed in the context of Coulomb failure stress (CFS) evolution since 1812 in southern California. We systematically considered circular regions and equal-area annuli centered on the epicenter of the Landers earthquake. Frequency-magnitude relationships for two 10-year periods before and two 5-year periods around the Landers event are compared. Only events with magnitude, M ≥ 4.0 are included; aftershocks are removed. For the larger circular regions with radii of 140 to 160 km, the rate and slope of the frequency-magnitude distribution for moderate-size events just before the mainshock appear to be anomalous compared to those for either the preceding or subsequent periods. For areas closer to the 1992 epicenter, however, the number of events is few, and the differences in the distributions are less obvious. When we examined the seismic activity in annuli of equal area, however, the largest changes occurred about 150 km from the epicenter of the mainshock, not closer as would be expected for a precursor to the Landers event. We also derive an index value to better quantify differences in the frequency of occurrence of moderate-size events as a function of time. The index value and the frequency-magnitude distribution show similar spatial dependence. Since 1812 a large region near Landers has moved closer to failure in terms of changes in CFS for faults of San Andreas type. These changes, however, are dominated by coseismic changes associated with the 1812 and 1857 earthquakes and by tectonic stress buildup related to the San Andreas fault, not by stress buildup associated with the Landers faults themselves, which are characterized by very slow long-term displacements. Hence, the most pronounced changes in the frequency of moderate-size earthquakes before 1992 do not appear to be related to stress buildup to the Landers sequence itself. They, along with the Landers sequence, may be indicative of a broad region that is approaching a high stress state prior to an eventual future great earthquake. The failure to find a pronounced increase in moderate-size shocks close in to Landers is in accord with the idea that such increases on a timescale of years to decades are associated with the regional buildup of stress to large earthquake along faults of high (not low) long-term slip rates.