What may be gained by applying cost-benefit analysis (‘CBA’ for short) to methodological problems? It is claimed that the perspective of CBA may help the researcher to see what sorts of questions he should take into account when dealing with particular methodological problems. This claim is supported by applying CBA-thinking to two standard problems of the philosophy of science. In section 1 the concepts of “economic approach” and of “methodology” are clarified. Section 2 is devoted to the problem of the empirical base. In the practice of research the handling of basic statements seldom, if ever, constitutes a problem, and no conscious decision is involved. In the methodological reconstruction of that practice the problem of the empirical ‘base’ is viewed as an investment problem: whether or not to invest time and effort into processing a particular basic statement into a falsifying hypothesis for the theory we wish to test. The valuation of the costs of rejecting or, as the case may be, of defending a basic statement, are objective. Section 3 deals with the problem of theory choice. In basic science the issue is not one of acceptance or rejection of a single theory, but rather of theory preference. Both the rational response to a falsification and rational theory preference are governed by CBA-considerations. The option for one of two competing theories is based upon a CBA where the valuation of benefits and costs is objective. Theory change is an objective process, at least in those fields where theorizing is closely monitored by empirical testing: in the long run the better theory drives out the less good theory. The costs of defending a theory that is less good than its competitor are mainly opportunity costs, epistemic resources forgone.