This question is the subject of a lively debate among professional historians (non-professionals are also pitching in, but I'd rather not discuss their contributions at this stage, as per my impression they range from thought-stimulating cherry-picking to outright hackery).
There is a recent (2009) review paper by Robert R. Dykstra: Quantifying the Wild West: The Problematic Statistics of Frontier Violence. Alas, I cannot access it now.
The first page of another article by Dykstra is available for free and I'll quote a bit from it to indicate some points:
As an expression of this, it is now widely believed that the frontier
West experienced interpersonal homicide of Homeric proportions. Once
the intellectual property of film director Sam Peckinpah and his
imitators, this conviction overtook western historians in the 1980s.
[...]
Dissenters from the reigning paradigm, although few, have occasionally
been heard. Thomas M. Marshal, Lynn I. Perrigo, and Michael N. Canlis
contested the portrait of violent, anarchic frontier mining camps
drawn by Bret Harte, Mark Twain, and other writers. None in early
Gilpen County, Colorado, said Perrigo, resembled "a traditional 'Wild
West' settlement, with each man a law unto himself." This
reviewercontributed the information that Dodge City, Abilene, and the
other fabled Kansas cattle towns were only intermittently violent and
hardly lawless; they averaged only one and a half adult homicides per
cattle-trading season. Harry H. Anderson revealed that literally
lawless Deadwood, South Dakota experienced only four killings in its
notorious first year. Frank Richard Prassel concluded from his survey
of frontier law enforcement that a westerner "probably enjoyed greater
security in both person and property than did his contemporary in the
urban centers of the East." W. Eugene Hollon agreed, contending that
the frontier "was a far more civilized, more peaceful and safer place
than American society is today."
As a professional should, Dykstra also warns from committing the fallacy of small numbers.
I also found in a Texan blog an example, apparently drawn from the first-mentioned paper:
Dykstra believes that "the fallacy of small numbers," which results in
over-inflated generalizations based on statistics from small
localities, is the unresolved weakness on the violent West side of the
debate. Thus, based on a single homicide in 1880, Dodge City,
population 996, had a murder rate three times that of Miami, in 1980,
which saw 515 murders among a population of 1.57 million. But was
Dodge City more violent than Miami?
So the jury is still out, as far as I can tell.