We evaluate the Friedman rule for optimal monetary policy in a laboratory economy based on Lagos-Wright (Journal of Economic Theory 145 (2010), 1508. We explore two implementations of Friedman's rule: one involving deflationary monetary policy and another where interest is paid on money. We compare the welfare consequences of the Friedman rule with two other policies: a constant money supply regime and a regime where the money supply grows at a constant k%. Counter to theory, we find that the Friedman rule is not welfare-improving, performing no better than the constant money regime. By one welfare measure, the k% money growth rate regime performs best.