Contralateral masking occurs when the threshold of a signal in one ear is elevated by the presence of a masker in the other, contralateral ear. The classic data and theory on contralateral masking were provided by Zwislocki [J. Acoust. Soc. Am. 52, 644-659 (1972)] who observed a 3- to 18-dB threshold shift (masking) for a gated pure-tone signal in one ear when a gated pure-tone masker was presented via insert earphones to the other ear. Zwislocki referred to this phenomenon as "central masking." Here, using two psychophysical methods (Yes-No; two-interval forced-choice), Zwislocki's original results, obtained with other psychophysical methods, were successfully replicated. Similar results using several psychophysical methods suggest that contralateral masking is indicative of a sensory phenomenon rather than observer bias and other response proclivities. In a second experiment, psychophysical tuning curves were obtained using either an ipsilateral masker or a contralateral masker. Tuning curves obtained with a contralateral masker had steeper slopes on both the low- and high-frequency sides than tuning curves obtained with an ipsilateral masker. Thus, although substantially smaller in effect than ipsilateral masking, contralateral masking is more sharply tuned. The sharp tuning of contralateral masking reflects a greater compression of the input/output functions for contralateral masking than for ipsilateral masking. The closest correspondence between the tuning curves reported here for contralateral masking and those predicted by Zwislocki's theory and data (on central masking) occurred for tuning curves where the ratio of driven activity to spontaneous activity was about six. A remaining issue is the role, if any, of the efferent auditory system, especially the olivocochlear bundle, in threshold shifts measured using the Zwislocki (central masking) paradigm.