Understanding the relationships between morphological disparity and environment, geography, and scale require examination at the local level. Even with disparity metrics that are inherently sample size independent, the nature of rare species and the segregation of common and rare species within morphospace can create substantial sampling issues. Eight well-sampled, Late Ordovician crinoid assemblages were examined for potential biases in the study of local disparity. Disparity is based on the ordination of discrete characters. The rare and common species within these assemblages contributed equally to disparity. In spite of this pattern, rare species in some localities occupy a different area of morphospace, causing disparity to vary greatly with sampling intensity. Morphological rarefaction based on the number of specimens shows that disparity weighted by abundance is constant past a sample size of approximately 30 individuals. This metric is dependent on the evenness within an assemblage as well as the abundance within subgroups in morphospace. Disparity weighted according to abundance gives a view of the functional disparity of an assemblage, which is more applicable in studies of local disparity, though unweighted disparity is still preferred in regional-scale studies and in investigations of morphospace filling through a clade's history.