Three hearing protection devices [Siebe Norton Com‐fit earplug, Flents Silenta model 080 earmuff, and E‐A‐R■foam earplug] were compared for their effectiveness in preventing both daily temporary threshold shifts (TTS) and threshold variability between sequential annual audiograms for workers in an industrial environment with a TWA of 107 dB. Com‐fit wearers showed significant TTS at 500–4000 Hz from the beginning to the end of the workday. Silenta wearers displayed significant TTS at 500 Hz and nonsignificant improvement at 6000 Hz. E‐A‐R wearers showed significant TTS at 1000–2000 Hz and nonsignificant improvement at 3000–6000 Hz. Statistical analysis indicated that the protection provided by the E‐A‐R plug was significantly better than that of the Com‐fit plug at 500–4000 Hz, and significantly better than the protection of the Silenta muff at 500, 3000, and 4000 Hz. Employees’ last four annual audiograms prior to the start of the TTS study were evaluated to compare the wearer groups’ mean values of the %BWs statistic (sequential percent better or worse). Previous research indicates that the %BWs statistic for a properly protected population with past audiometric test experience will be less than 30%. The E‐A‐R group showed an acceptable %BWs value (26%), but the Silenta wearers (45%) and Com‐fit wearers (53%) showed excess threshold variability in threshold measurements from year to year.