It's a data set that supports some degree of trend analysis... that's about it. Some of the data is interesting, some is eye-opening such as the DUI numbers but none of it is conclusive of anything in and of-itself short of the demographic data on the decedents and the like.

Debating the numbers, collection methods and the like is both interesting and perhaps entertaining, but rarely does it stimulate thought when the ones debating are deeply rooted in given position or set of presumptions and biases.

If a certain number piques your interest, by all means explore it...

My only point in offering a subequent response to some of the early posts was to make sure those looking at the data remind mindful that a lot of the cyclists contained in these numbers are not what the cycling community -- particularly the ones who have access to a keyboard and who frequent internet discussion boards -- usually think of first when they hear a cyclist has been killed or look at this type of widely reported fatality data... noting that it's what everyone else will pick-up and feed off of, e.g., the IIHS (yes, it's simply their analysis of what the NHTSA reported in June).