Tests of the noise excluding properties of the TDH,50P/MX‐41AR, the Audiocup, and the ER‐3A insert earphone were performed in a diffuse‐field facility complying with ANSI S12.6‐1984. Data on attenuation were obtained monaurally with the nontest ear plugged and muffed. Results generally agreed well with previously reported measurements. A broadband noise, shaped by a multifilter, allowed a direct test of the ANSI S3.1‐1977 permissible background noise levels for testing to audiometric zero under TDH‐39/MX‐41AR headphones. This “ANSI noise” raised the average thresholds of 15 normal‐hearing test subjects by 3 to 5 dB at the 500‐ to 4000‐Hz octave frequencies. With a noise shaped to the less stringent OSHA‐1983 regulation, average thresholds were raised 9 to 17 dB. Introduction of an “ENT office noise” with 55 dB(A) overall level raised average thresholds at those frequencies by 11, 5, 1, and 0 dB with the Audiocups and less than 2 dB with the recommended fully (“deeply”) inserted ER‐3A eartips. Measured threshold elevations agreed closely with predictions based on a critical ratio calculation utilizing actual sound field noise levels and measured attenuations. A conservative rule of thumb for testing to audiometric zero with the ER‐3A earphone appears to be an office noise level of 45 dB(A) or less.