Issues in the accuracy of the tornado climatology of the United States prior to 1980 have been well documented (Kelly et al. 1978; Doswell and Burgess 1988). While some of these issues center on low population (few observers to note tornado occurrences) and poor understanding of the definition of a tornado, an additional factor was at play in California before the 1990s. This centered on the fact that meteorologists of that time were accustomed to “pattern recognition” in anticipating occurrences of tornadic thunderstorms (rather than on the ingredients that produce them). Since the synoptic patterns associated with California tornadoes do not resemble those associated with the quintessential Great Plains tornadic thunderstorm, there was a considerable bias that tornadic thunderstorms could not occur in California. The authors believe, because in the minds of California meteorologists, California tornadoes were weak, non-supercellular and seldom had long track lengths and because of this tornado reports were often dismissed as wind reports, and meteorologists frequently were slow to survey the tornadoes that did occur. Two of the most damaging tornadoes in the California record were the 1951 Sunnyvale F2 tornado and 1983 Los Angeles Convention Center F2 tornado. There is evidence that the former was not surveyed in a timely manner, and that the latter was given a mistaken rating. The authors have uncovered many photographs and eyewitness accounts that suggest the Sunnyvale tornado produced at least F3 damage, and was one of a swarm of tornadoes that struck the San Francisco Bay Region that day. There is similar evidence in the newspaper archives that the Convention Center tornado may have produced borderline F4 damage. Thus, due to the lack of understanding of tornadic environments by California meteorologists at the time, these events were documented poorly meteorologically and the Fujita scale rating given to each was probably underestimated. A case study of synoptic and sub synoptic conditions for both of these events is provided that suggests that a reanalysis of the damage reports and surveys might suggest a reclassification of the F or EF rating for both events.