This study describes the effects of gender and age on peer acceptance in primary-grade, multiage classrooms at Edith Bowen Laboratory School at Utah State University. The population described consisted of six multiage classrooms composed of male and female students from 6 to 8 years old. The classrooms were approximately balanced by gender and age. Students spent the entire day and received all instruction in the multiage setting. Students completed "Work With" and "Play With" sociometric rating-scale instruments. For both instruments, results revealed the existence of "gender cleavage"--both genders preferred work and play partners of their own gender. In general, age accounted for more variance in peer acceptance scores for older students than younger students. However, eta-squared statistics demonstated that except for 8-year-old males, gender accounted for much more of the variance in peer acceptance scores than age. The correlation coefficient for paired peer acceptance scores for each student on the two instruments was .94. These results support the conclusion that students did not differentiate responses based on "Work With" and "Play With" criteria . Test-retest reliabilities for the "Work With" and "Play With" instruments with a 1-week testing interval were .94 and .92, respectively.