Purpose: The ability to detect a tone added to a random masker improves when a preview of the masker is provided. In 2 experiments, the authors explored the role that perceptual organization plays in this release from masking. Method: Detection thresholds were measured in informational masking studies. The maskers were drawn at random prior to each trial. Masker or signal-plus-masker precursors preceded the detection interval, and the time between the precursor and the detection interval was systematically altered. In Experiment 1, the signal frequency was either fixed or random. In Experiment 2, the random masker was composed of harmonics of a common fundamental frequency (F [subscript 0]), and the randomly chosen signal frequency was either harmonically related to, or mistuned from, the masker's F [subscript 0]. Results: For a masker precursor, the release from informational masking withstood longer precursor-detection interval delays (a) when the signal frequency was fixed versus random and (b) when the signal was mistuned relative to a harmonic of the masker's F [subscript 0]. Conclusion: These results suggest that listeners' ability to attend to the signal may contribute to the long-lived release from masking with a masker precursor.