# When does Pearson's chi-squared test fail?

In which cases can't I perform a Pearson's chi squared test because the outcome will be meaningless? Edit: Is it still valid even in the case the distribution does not allow mean and variance?

-
It sounds to me that there are both statistical and pragmatic (i.e., related to a research study) reasons where a $\chi^2$ test would be meaningless. Could you perhaps refine your criteria and tell us what application of Pearson's $\chi^2$ you are interested in? –  chl Jul 31 '12 at 10:21
In light of your edit, are you interested in testing whether an observed frequency distribution differs from a theoretical distribution? –  chl Jul 31 '12 at 12:43