In this paper we derive a new method for evaluating or approximating sums by means of complex integration. Our result is sufficiently general that it is applicable to a wide variety of functions. We consider examples that illustrate the power of the technique; our first example is an alternative derivation of the Euler-Maclaurin sum formula for the case in which the remainder term vanishes, and our other two examples show how our technique can be applied when the Euler-Maclaurin formula is not useful.