UCAC3 is supposed to be less prone to error than the USNO-A2.0 catalog. It has a lower limiting magnitude (16), so in some cases the A2 data would still be useful.
The USNO-A2.0 data is known for its high photometric error. I'm not certain how it compares in that respect to UCAC3. Also, USNO-A2.0 does not have proper motions, and USNO-B1 has significantly higher estimated error in positional measurement than UCAC3.
USNO-B1 is still the largest & deepest data set, but rather hard to get your hands on if you need the whole database
One of the goals of UCAC3 was to be complete to magnitude 16, which is good for most of my usage. Beyond that you can still query USNO-B1 online for the region of interest.
Regards,
Eric