The basis for detection in those interaural conditions that lead to a masking‐level difference (MLD) is presumed to be quite different from that in the non‐MLD conditions. This experiment examined the effect upon detectability when listeners were uncertain whether an MLD or a non‐MLD condition was to be presented. Detectability was measured at two values of signal energy for each of two interaural conditions, N0‐S0 and Nπ‐S0. The “YES‐NO” psychophysical method was used. On some blocks of trials, the blocks with “no uncertainty,” only one interaural condition was presented. On other blocks of trials, the blocks “with uncertainty,” N0‐S0 signals were mixed with equally detectable Nπ‐S0 signals. The listeners were informed which type of block was to be presented. There was essentially no difference between the data obtained with uncertainty and those obtained with no uncertainty. A supplementary experiment showed that on the blocks with uncertainty the listeners were quite poor at identifying the type of noise presented on a trial.