The first obvious example is base 2, what computers use. Computers really don't have a way of storing "part" of a value, so they deal with litteral "on" and "off" values, which are 1 and 0 respectively. Base 2 is very helpful in this arena because it allows computers to do a lot of math using just on and off signals. (Base 2 is often called "binary".)
A natural extention of base 2 is base 16 (since 2^4 = 16). This base (in computer terms called "hexadecimal") is used to abbreviate base 2 numbers since not everyone wants to have to write out all the 1s and 0s, so every hexadecimal digit corresponds to 4 binary digits.
You'll occationally see base 8 in use (2^3, go figure), but it's fairly uncommon. I believe it's most practical use is in certain biology aspects.
There is also the natural base, e, which serves as a fundamental base number that makes recursive appearances in calculus and aids in making some calculations simpler. 'e' has no rational decimal approximation (like pi), it's equivilent about 2.78...
Overall, though, the number base you choose to use for something is up to you, and it's all about convenience. Realistically, it doesn't matter what base you use, it's still the same number. The only difference is how you say it, and how easily you say it. Some bases may be faster for computer math calculations, while others may be more consise, while others may aid in solving math problems; it's just about what helps.