Some other good sources:
JC Bradbury's studies (click on the Age and Hitting Links)
Baseball Notebook's David Luciani's study
http://www.baseballnotebook.com/essay.a ... %20Hitting
Keith Woolner's look
A brief comment on Bill James' original 1982 study can be found here:
http://baseballanalysts.com/archives/20 ... rom_17.php
at the bottom of the page
Ken Cherven has done some recent work on this and presented it at SABR 2005 in August
http://220.127.116.11/search?q=cache:P0 ... ball&hl=en
I can't seem to get the powerpoint to work, so the link is an HTML version of his slides.
If you really like fancy stats, Jim Albert's study is for you
And Medline even shows a study
http://www.ncbi.nlm.nih.gov/entrez/quer ... query_hl=3
Given Roy Fair's skills, it will be interesting to see this full study when it is published...
http://ideas.repec.org/p/cwl/cwldpp/153 ... /2005D.PDF
It also has a wonderful review of possible juicing based on looking at players who out-performed what his model predicted.
In general, I would say that people often misinterpret the early studies, which really never pinpointed a single age as THE peak. It's much more accurate to say that they indicated that the vast majority of players peaked between ages 25 and 29, with 27 the most common peak, by a whisker. Furthermore, all the studies show that peak age differs for different statistics: batting average and SBs tend to peak before 27, power stats like homers and doubles peak right around 27, and stats like walks and OBP tend to peak after age 30.
The more recent studies, like Bradbury's and Cherven's suggest two important things. First, the peak seems to be shifting a little later, with maybe 26 to 31 as a better peak window for today's players. They also suggest that players are declining more slowly after the peak. There's lots of speculation about causes, including misreported ages in early datasets, steroids, better training, and a bunch of others.