For effectively annotated phenotypes for mouse, a number of detailed phenot
ypic classification systems are needed. The rapidly increasing number of ph
enotypically described characteristics of both normal and mutant mice are p
roviding a rich data set for comparison and analysis. However, we cannot re
ly on text descriptions that are subject to the word-usage style of the: wr
iter if we are to do large-scale comparative analysis of traits and disease
s. The rationale for developing vocabularies and examples of several vocabu
laries being developed are described. Finally, the critical nature of commu
nity participation in both building and applying phenotype vocabularies is
discussed.