Is mathematical ability a bell curve?

My own anecdotal experience has been that women are rare in elite math environments, but don't perform worse than the men. That would be consistent with a fat-tailed rather than normal distribution, and also with higher computed variance among women.

Also anecdotal, but it seems that when people come from an education system that privileges math (like Europe or Asia as opposed to the US) the proportion of women who pursue math is higher. In other words, when you can get as much social status by being a poly sci major as a math major, women tend not to do math, but when math is very clearly ranked as the "top" or "most competitive" option throughout most of your educational life, women are much more likely to pursue it.