We applied the Bakun and Wentworth (1997) method to the 26 March 1812 Venezuelan earthquake. Previous studies have shown that it had at least two large subevents, one associated with the Boconó fault and an intensity center between Barquisimeto and San Felipe, the other associated with the San Sebastián fault near Caracas. We used the method with two modifications. First, we introduced intensity intervals to account for the uncertainties associated with historical information. Second, we incorporated a correction that considers the fault’s strike, but not its actual position. These modifications were tested with three earthquakes: the 1989 Loma Prieta; a 1980 Mw 5.2 event in Táchira, Venezuela; and the 1967 earthquake in Caracas, Mw 6.6. The third one was important for the interpretation of the 1812 Caracas subevent. For the Barquisimeto–San Felipe subevent we obtained MWI 7.4±0.35, and an intensity center 10.20° N, 69.95° W that lies between Barquisimeto and San Felipe. Our preferred rupture length was between 90 and 100 km. The Caracas subevent had MWI 7.1±0.33, with an intensity center 10.60° N, 67.10° W, close to the Venezuelan north coast. The expected rupture length was on the order of 70 km.