If you are interested in a particular game, and the number of pitches isn't listed in the box score, try going to the opponent's website and reading their box score. It is my belief that number of pitches is recorded in college automated scoring systems, but not all teams like to publish them.
When neither school publishes the counts, then it is possible to estimate the number of pitches by empirical formula. Several have been published, and, on average, they of course provide an accurate estimate of the number of pitches. However, there is some variance between calculated and actual pitch counts. In my opinion, using calculated pitch counts to identify unusually high counts is fraught with inaccuracy, because, for example, those games which have suspiciously number of high batters faced are likely the ones that had an unusually low number of pitchers per batter faced. At least that's what I found last year when looking at compete games for which the number of pitches was published. The formula tended to overestimate (substantially) for complete games.
One commonly used formula is NP=3.3PA + 1.5SO + 2.2BB (
http://www.tangotiger.net/pitchCounts.html)
Boyd Nation has a much more complicated one.
EPC formula