All data are from the official NWS tornado database (for tornadoes 1950-1995), and from Storm Data (1996-present). The 1996-present data was copied manually and therefore may have errors. The only other modifications made were similar to those described in my Corrections to the Historic Tornado Database Technical Memorandum, when apparent. This does not infer that data on my lists are necessarily "correct".
For reasons of space and upkeep, I did not want to include the entire list of tornadoes in the NWS tornado database. But, I wanted to include more than just strong/violent (F2+) tornadoes since this list can be quite biased against largely rural areas. (It's difficult to obtain an F2+ rating when there are few targets to hit). So I arbitrarily decided to also include tornadoes with a length of 5 or more miles, and tornadoes with widths of 400 or more yards. For a more complete list of tornadoes in a particular county, you can refer to various local National Weather Service Forecast Office pages (such as the Norman NWS tornado data by county page that I also maintain) or the National Climatic Data Center's online storm events database (although realize that the NCDC database is a county-based list, and continuity with county border-crossing tornadoes is difficult).
At this point, I have also started the lists at 1980. Documentation in recent years has become better in general, but there still are tornadoes in this list described as "intermittent," which indicate various touchdown points, but counted as a single tornado. More recent events such as this would often be seperated into individual members of a tornado family.
Not all data has been cross-checked, especially those that crossed state lines. And in Storm Data, unless it has been described in the text, it is difficult to match tornadoes that cross county boundaries. These tornadoes have been combined into a single tornado event when I have noticed the text, but there may be occasions where it was not noticed or not described in the Storm Data text.
I'm sure that there are more caveats that I can't think of now.