Two experiments on forward masking are described. Probe threshold were measured using a two‐interval forced‐choice technique. In experiment I the masked threshold for a 1‐kHz probe was measured as a function of masker level for three maskers; a 1‐kHz tone, a 1/3‐, and a 1/1‐octave noise band centered at 1 kHz. At medium and high masker levels, maskers equated for energy in the critical band around 1 kHz did not produce equal amounts of masking. The tone was the most effective masker, and the 1/1‐octave noise the least. In experiment II the threshold for a 1‐kHz probe was measured for a 1‐kHz masker at levels ranging from 26 to 76 dB SPL, and for maskers ranging from 800 to 1100 Hz at levels of either 76 or 61 dB SPL. The effect of adding to each of these a 1.2‐kHz tone at 81 dB SPL was also determined. The 1.2‐kHz tone produced a significant reduction in probe threshold for masker frequencies within 50 Hz of that of the probe, but had little effect for maskers outside this range. The results are discussed in relation both to lateral suppression and to cues arising from qualitative differences between masker and probe.