The 1906 San Francisco earthquake was caused by a rupture on the , a continental that forms part of the boundary between the and the . This runs the length of California from the in the south to to the north, a distance of about 800 miles (1,300 km). The earthquake ruptured the northern third of the fault for a distance of 296 miles (477 km).
A strong preceded the mainshock by about 20 to 25 seconds. The strong shaking of the main shock lasted about 42 seconds. The shaking intensity as described on the reached 8 in San Francisco and up to 9 in areas to the north like where destruction was devastating.
There were decades of minor earthquakes – more than at any other time in the historical record for northern California – before the 1906 quake. Widely previously interpreted as precursory activity to the 1906 earthquake, they have been found to have a strong seasonal pattern and have been postulated to be due to large seasonal sediment loads in coastal bays that overlie faults as a result of the erosion caused by "" in the later years of the .
As damaging as the earthquake and its were, the fires that burned out of control afterward were much more destructive. It has been estimated that up to 90% of the total destruction was the result of the subsequent fires. Over 30 fires, caused by ruptured gas mains, destroyed approximately 25,000 buildings on 490 city blocks. Worst of all, many were started when , untrained in the use of , attempted to demolish buildings to create , which resulted in the destruction of more than 50% of the buildings that would have otherwise survived. The city's Fire Chief, Dennis T. Sullivan, who would have been responsible, had died in the initial quake. The dynamited buildings themselves often caught fire. In all, the fires burned for four days and nights.
Due to a widespread practice by to indemnify San Francisco properties from fire, but not earthquake damage, most of the destruction in the city was blamed on the fires. Some property owners deliberately set fire to damaged properties, in order to claim them on their insurance; this ultimately served no purpose, as wealthier citizens of the city shouldered the costs of repairing an estimated 80% of the city. Capt. Leonard D. Wildman of the reported that he "was stopped by a fireman who told me that people in that neighborhood were firing their houses...they were told that they would not get their insurance on buildings damaged by the earthquake unless they were damaged by fire."
For a number of years, the epicenter of the quake was assumed to be near the town of , in the area of , because of evidence of the degree of local earth displacement. In the 1960s, a seismologist at proposed that the epicenter was more likely offshore of San Francisco, to the northwest of the . However, the most recent analysis by the (USGS) shows that the most likely epicenter was very near on the coast of , an adjacent suburb just south of San Francisco. An offshore epicenter is supported by the occurrence of a local recorded by a tidal gauge at the ; the wave had an amplitude of approximately 3 in and an approximate period of 40–45 minutes.
The most important characteristic of the shaking intensity noted in Lawson's (1908) report was the clear of intensity with underlying geologic conditions. Areas situated in -filled valleys sustained stronger shaking than nearby bedrock sites, and the strongest shaking occurred in areas of Bay where landfill failed in the earthquake (). Modern seismic-zonation practice accounts for the differences in seismic hazard posed by varying geologic conditions.