Generally, the closer to expiration you get, the more pronounced the smile becomes. This is some German Bund data I believe.
Chilipalmer, Panzerman has posted a great 3D chart. The front month options generally have more skew than the back month, thus the 'volatility smile'. But you need to also remember that the skew itself changes and, as Panzerman pointed out, is affected by the passage of time (front month versus back months) and other changes in market conditions such as a change in iv or a change in price of the underlying. One of the other members also mentioned the severity of the skew, i.e. goog has more specific risk than an index and thus more severe vol smile gradients. There is a good section in Natenberg on these skews. Also, if you check out TOS there is a tab for 'volatility analysis' (under 'analyze') and it supports what panzerman said. One more thing to remember is that these 'smiles' are all derived from data, some of which may not be terribly accurate or up to date. Iow some of the vol inputs may come from options that are not very liquid and thus have large bid/ask spreads (for example options that are far otm or deep itm). The question is 'how can you use these skews to your advantage?' Again, it comes down to predicting what volatility is going to do, either in the skew or in the specific option you want to trade. And, at the same time, guessing how the change in vol will be offset by a change in the price of the underlying. Cheers db