Postfix operators, namely ++ and –, are less efficient than their prefix counterparts because the compiler usually creates a temporary copy of the object before applying the operator to an object. When using fundamental types such as int or char, most compilers can optimize the code and eliminate the performance overhead associated with postfix operators. Therefore, there’s no real difference between the the following expressions:

  for (int j=0; j

However, if we use a user-defined type instead of a fundamental type, most compilers won't be able to optimize the postfix version. In the following example,

  string s = "hello"; string::iterator si; for (si= s.begin(); si

Chances are high that the prefix version of the for-loop is more efficient than its postfix counterpart because si is an iterator, not a fundamental type. For this reason, always use prefix operators unless you truly need to use the postfix version.