The study, involving 15 emotionally (EH), hearing (HH), or physically (PH) handicapped children in segregated preschool special education classrooms, 15 mainstreamed EH, HH, or PH preschoolers, and 15 nonhandicapped preschool peers, analyzed interactions of both handicapped and nonhandicapped children, ecological information on each classroom setting, and nutritional data obtained from parents of each child. Five hypotheses were proposed: (1) in general, the level of play of handicapped preschoolers will be lower than the nonhandicapped peers; (2) there will be a larger number of adult initiated interactions by teachers/aides in the segregated settings than in the mainstreamed settings; (3) there will be more solitary and parallel play in segregated settings than in mainstreamed settings; (4) depending on the category of handicap, the predominant category of play interactions will be varied in the mainstreamed and segregated settings; and (5) Ss without breakfast will exhibit more solitary play than those Ss who ate breakfast. Among findings were that although the nonhandicapped Ss had higher frequency scores in all categories of play, in the practice play category, mainstreamed handicapped preschoolers scored lower in only one instance; and that in the segregated settings during the observations there appeared to be more adult interaction during the free play periods than in the mainstreamed settings. (SBH)