This sounds like half the question. The quickest figure out the maximum value of an unsigned number on a given system is to take the bitwise complement of 0: unsigned long n = ~0;