The acquisition of the copula and auxiliary "be" forms of "am,""is," and "are" was studied with 16 children. Spontaneous social speech samples were obtained from each child. One child's speech was sampled for approximately one hour per week from age 2;5 through 4;0, and for one-half hour per week from age 4;1 to 5;0. A second child's speech was sampled for approximately one hour per week from age 1;11 to 3;6. A weekly one hour sample of speech was obtained from the other children for six consecutive weeks. The speech samples were analyzed for omissions, incorrect uses, and correct uses of "be" allomorphs in specific sentential frameworks (e.g., copula "are" following "they" and preceding a prepositional phrase in the declarative). Results include the following: (1) In the acquisition of a "be" allomorph, the predominant pattern was one in which children learn to use the allomorph in quite specific contexts. This was true both for the child's initial uses of the form and the subsequent stable acquisition of the form. (2) Children do not seem to learn all members of the copula or the auxiliary "be" class simultaneously; in addition to varying patterns of initial uses, differences were found in regard to age of final acquisition (i.e., 90 percent correct usage in obligatory contexts). (3) Copula "be" forms tended to be earlier stable acquisitions than auxiliary "be" forms. (4) Most "be" forms were acquired (90 percent criterion) in declarative contexts before yes-no question contexts and "be" forms also tended to be acquired (90 percent criterion) in yes-no question contexts before "wh" contexts. (SW)