I agree Nutcase that being in the same genus doesn't automatically mean they can interbreed, just that it's physically more viable. There can still be behaviorally differances that prevent it (as in the case with Satanoperca jurupari and S. leucosticta).
Usually a lion and a tiger won't interbreed either, but crosses have been forced bred before and they do create a rearable young. Efforts to try to cross genus to a cheetah have failed though. Wolves and coyotes usually won't cross either, but it has happened. The more closely they are related, the higher the percentage the viablity of the young seem to have.
It is anecdotal perhaps, but seems to bear through with mammals. Fish are a differant beast though. Convicts can cross with many central americans true, but seem to have the easiest time with other members of their closesly related groups (Cryptoheros and Archocentrus) than they do with with say a Vieja or Parachromis.
Of course, I'm not forgetting that in nature failure to cross breed helps determine species ... even with convicts. But we all know what happens in a tank can be quite differant.