The neo-Fisherian effect typically refers to the short-run increase in inflation associated with a permanent increase in the nominal interest rate. This positive comovement between the two variables is commonly viewed — and empirically identified — as being conditional on permanent monetary shocks, which are often interpreted as permanent shifts in the inflation target. Such a view, however, implies that inflation and the nominal interest rate share a common stochastic trend, a property that is hardly supported by the data, especially during episodes of stable inflation. Moreover, in countries that have adopted formal inflation targeting, changes in the inflation target occur very infrequently, if at all, calling into question the interpretation of inflation target shocks identified within standard time-series models based on quarterly data. In this paper, we propose a novel empirical strategy to detect the neo-Fisherian effect, which we apply to U.S. data. Our procedure relaxes the commonly used identifying restriction that inflation and the nominal interest rate are cointegrated, and, more importantly, is agnostic about the nature of the shock that gives rise to a neo-Fisherian effect. We find that the identified shock has no permanent effect on the nominal interest rate or inflation, but moves them in the same direction for a number of quarters. It also accounts for the bulk of their variability at any given forecasting horizon, while explaining a non-negligible fraction of output fluctuations at business-cycle frequencies. Using Bayesian techniques, we show that the data favors the interpretation of the identified shock as a liquidity preference shock rather than an inflation target shock.