The insertion losses of four pairs of earmuffs, including one noise-excluding headset, were measured in one-third octave bands in a diffuse broadband noise field using a head-like acoustic test fixture. The acoustic test fixture contained realistic ear simulators with microphones at the eardrum positions. The insertion losses were measured (i) with the earmuffs on their own, (ii) with the earmuffs worn over an anti-flash hood, and (iii) for one earmuff, with the earmuff worn under the hood. The other three earmuffs could not be fitted under the hood. The insertion loss of the anti-flash hood on its own was also measured. Wearing an anti-flash hood under the earmuffs greatly reduced the protection against noise, by 20–23 dB at high frequencies, by 17–20 dB at middle frequencies, by 12–16 dB at low frequencies, and by 16–20 dB overall. Only one earmuff was slim enough to fit under an anti-flash hood. Wearing an anti-flash hood over this earmuff had only a marginal effect on the earmuff insertion loss, of the order of 1 dB. If anti-flash hoods could be designed to fit over other types of earmuffs and headsets, the protection of these earmuffs and headsets would be virtually maintained.