Just-noticeable interaural time differences were measured for low-frequency pure tones, high-frequency sinusoidally amplitude-modulated (SAM) tones, and high-frequency transposed stimuli, at multiple levels with or without a spectrally notched diotic noise to prevent spread of excitation. Performance with transposed stimuli and pure tones was similar in quiet; however, in noise, performance was poorer for transposed stimuli than for pure tones. Performance with SAM tones was always poorest. In all conditions, performance improved slightly with increasing level. The results suggest that the equivalence postulated between transposed stimuli and pure tones is not valid in the presence of a spectrally notched background noise.