Critics of creativity tests, some of which are used widely in determining student eligibility for special educational opportunities such as gifted programs, claim that scores on these tests can be improved with specific training on the scoring components of the tests. However, we could find no empirical evidence supporting this claim. So, we tested this by comparing pre- and post-scores on the Torrance Tests of Creative Thinking (TTCT) of three groups of children: (a) those who received specific training; (b) those who received no training, and (c) those who received general creativity skills training. Results indicated that training on the criteria of the TTCT did not improve scores more than general creativity skills training or no training at all. Scores for all three groups improved similarly on retesting, but only slightly.