DD took it in second grade but was given the third grade version and it was timed. The CogAT website recommends this to school districts (giving kids one grade level ahead) to find the most gifted kids. Not sure what the reasoning is, even if they are given the right level grade there should be a percentile associated with it. Not everyone is going to score above the 99th percentile. But maybe in some locations there are a lot of kids who score high. The CogAT is not an IQ test, the skills on it can be taught. I have a friend whose DD took the CogAT last year and scored something like a 118 composite. They prepped her and she took it again the next year and her score went up to 152. Didn't know it was possible to score that high, but that's what the mom said.
My DD didn't do well because of the fact that it is timed and she doesn't do well on timed tests because of her ADHD. So if a school is trying to eliminate twice exceptional kids from gifted programming, the CogAT is a good test to give. DD's WISC General Ability Index (composite of non-verbal and verbal) is well above the 99th percentile but only her CogAT verbal section showed a good score, because that is the section she finished in the time limit.
Really hate the CogAT. The untimed versions are probably more fair than timed versions just because timed tests will pretty much eliminate any kid that is slow (for whatever reason, could be that they are careful or perfectionistic), even if they do have very strong reasoning ability.
