Sampling is an important issue in the analysis and interpretation of f
aunal remains from archaeological sites. Use of various screen sizes i
nfluences the representativeness and sample size of remains collected
during excavation. In this study, the effects of screen size on measur
es of diversity were examined using faunal remains excavated from Nu'a
lolo Kai, Kauai Island, Hawaii. Variation in richness, the number of t
axa present, and evenness, the distribution of abundance values, is sh
own to be related to sampling biases produced by different recovery me
thods. Consequently, interpretations of prehistoric human subsistence
from faunal remains recovered by the larger screen sizes are questiona
ble. The use of smaller screen sites should be considered the more eff
ective approach to sampling a representative range of faunal remains.