This paper is primarily concerned with issues of symmetry–asymmetry in the joint dependence of binaural performance on interaural time delay and interaural amplitude ratio. The predictions of our model based on auditory‐nerve data are shown to be inadequate for available data showing asymmetries in interaural time discrimination. Measured detection thresholds for a 500‐Hz tone burst masked by wide‐band noise are reported for nonunity amplitude ratios and both polarities of delays in order to test for asymmetries analogous to those observed in the time discrimination case. The detection results show only small asymmetries and are not inconsistent with the model. A general discussion of our model and the implications of available empirical results leads to a suggested modification of the model. We show by an example that modifications of the type we are suggesting are capable of describing the detection data and at least the gross trends of the discrimination data.