This is an old original post, but it does need updating.
For a few months now, 23and me has had an initial cost of only $99, with a requirement for a minimum 1 year subscription for data updates at $9/month (charged monthly only after you results have been analyzed and posted). To me this is a pretty darn good price given what you get: autosomal DNA that links to all your lines; mitochondrial DNA to link to your maternal line; and (for males) Y DNA that links to your direct paternal line. Plus you get health related results.
The "raw data" you get can be viewed using several of their own tools, and their are several free services that you can use to try and get more discrete results. It appears to me that 23andme's user community is also much more helpful and knowledgeable than FTDNA users. (Sort of like RootsChat users being more knowledgeable and helpful than many other genealogy website users.)
I use both FTDNA and 23andme, so I son't think I have an overly biased opinion. Before testing with 23and me I signed up for a login with them and reviewed all their forums and information for over a month. FTDNA's autosomal testing is nearly 3 times as costly as that of 23andme. A benefit of doing FTDNA testing is that for strictly Y DNA testing ( a separate cost item) they use STRs rather than SNPs - see Wikipedia for definitions - although both do SNP testing for the autosomal DNA.
I believe that 23andme currently has a much larger database of autosomal testers, and you can add your data (for free) to one of their competitor's databases (DecodeMe [only relevant if you are interested in the health, rather than ancestry, bits]). There are strong rumors that FTDNA is going to allow 23andme customers to also add their data to the FTDNA database. This only makes good business sense.
I really wish that every serious family researcher that can afford it would do this autosomal testing. The findings tell you who else in the database is a likely 2nd, 3rd, 4th, 5th, or "more distant" cousin. It doesn't take too much imagination to see how a larger database could rather quickly lead to enabling us to break down those brick walls of the 1600s or 1700s.
Please look at their website.
Nick
