This study examined listeners' ability to process interaural temporal differences (ITDs) in one of two sequential sounds when the two differed in spectral content. A correlational analysis assessed weights given to ITDs of simulated source and echo pulses for echo delays of 8–128 ms for conditions in which responses were based on the source or echo, a 3000-Hz Gaussian (target) pulse. The other (distractor) pulse was spectrally centered at 1500, 2000, 3000, 4000, or 5000 Hz. Also measured were proportion correct and proportion of responses predicted from the weights. Regardless of whether the echo or source pulse served as the target, target weight, and proportion correct increased with increasing distractor frequency, consistent with low-frequency dominance [Divenyi, J. Acoust. Soc. Am. 91, 1078–1084 (1992)]. Effects of distractor frequency were observed at echo delays out to 128 ms when the source served as the target, but only out to 64 ms when the echo served as the target. At echo delays beyond 8 ms, recency effects were exhibited with higher proportions correct obtained for judgments based on the echo pulse than the source pulse.