Choice of the earmold for a particular individual is perhaps the most subjective of all decisions made by clinicians in hearing aid fitting. The four earmold designs used in this study were standard-long, standard-short, belled-vented, and cavernous. Conslusions were: 1) performance differences in earmolds of contrasting design can be assessed by discrimination scores with monosyllabic word lists of equal difficulty; 2) among the physical measures taken of the earmold-receiver systems, frequency-response overall range (HAIC) ranked the devices in agreement with group responses to a word identification task; and 3) these data offer a basis for endorsement of the standard-long and cavernous-short earmold types over the other two with their superiority evident in each of the discrimination categories of excellent, good, and poor.