"C" is a rating that specifies how much current the pack can support in a discharge. The "C" rating multiplied by the pack capacity will give you a number in Amps.

A specific example is a "20C" 4000mAHr pack. So the max discharge should be 20*4000 mA (ignore the "H") =80,000mA or 80 Amps.

I note that an actual 10C discharge would completely discharge the battery in 1/10 of an hour, or 6 minutes. In the previous example, that would be a 40 Amp continuous discharge.

"C" isn't universally defined by the Lipo vendors--sometimes it seems to be a marketing gimmick.

You basically should know that the main limit on a Lipo is how hot it gets during a discharge, and that is what sets the "C" limit. Recall that heating goes as the square of the current (in Amps) * the internal pack resistance. A lower internal resistance pack will have a higher C rating, so that is the physical meaning of "C".

You can stop here and not read the rest!

Another detail is that you can't calculate the heating with "average" current--the current we typically measure with our watt-meters and data recorders. So for an example here, if you are flying at 75% throttle on the ESC and measuring an average current of 40A, the actual current is 40A/0.75 or ~54 Amps for 75% of the time, and 0 Amps for 25% of the time. The heat produced during the off period is just 0 watts, but the heat during the on time is (40/0.75)² * R_int the internal battery resistance. However since this is only happening 75% of the time, the actual heating is just 40²/0.75 * R_int. Notice this is 4/3 (just 1/0.75) times  the heating you would have calculated using the average 40amps.

So looking at the 20C 4000mAHr pack, a 40A discharge would be a 10C discharge. However if you are pulling this at 75% throttle, you are actually heating up the battery pack as if you were making a 13.3C discharge. So partial throttle does exact a higher price on a pack than full throttle (where average amps=instantaneous amps).

Thanks Alan. We should be paying you a consulting fee for all this advice. I did read the whole thing, and may be able to calculate the expected heating of a given battery at a given current draw. If I understand you correctly, in general terms we want a higher C rating in a pack of a given size, to minimize heating and loss of efficiency, e.g. a 40C unit is preferred to a 20C unit rated at the same mAh? And partially to counteract the advertising puffery that may go along with the advertised C rating; buy a 40C cell to improve your chances that it really will perform as would a 30C cell if the advertised rating was real.

I'd say just enough so as not to use to much and not enough to get it to hot!  Mama bear said JUST right.

Alan, it is even worse, that curren going 75% of time out of the battery will go those other 25% of time back to he battery :-)  (via diodes in FET transistors) ... it will be something like back charging current. So theoretically that current in 75% of time must be even higher (to keep that average current 40A). Fortunately we have long leads from batery and large capacitors on ESC acting like a LC filter so it is not so bad in reality.