Based on Vygotsky's theoretical construct that play creates the zone of proximal development, this study was designed to examine the processes and outcomes involved in collaboration among dyads of 4-year-olds (matched for equal and unequal levels of play) in the context of constructive play with blocks. During the first phase of data collection, 100 4-year-olds were observed in naturalistic settings using the Play Observation Scale (Rubin, 1989). Play level, gender, and unfamiliarity were used to select 48 children to play with a peer in a laboratory setting. Play sessions were videotaped and coded for block play (Reifel & Greenfield, 1982), peer interaction (Rubin, 1989), and communication (Farver, 1992). Results of an overall multivariate analysis of variance (MANOVA) conducted for boys and girls found a significant interaction between treatment (play level) and gender. A follow-up MANOVA for girls was also significant. Subsequent univariate tests found significant differences in block play complexity of girls in treatment groups. A separate pairwise MANOVA found that less-skilled girls engage in more-complex play when paired with more-skilled peers. Block play complexity and communication contributed to the differences among the groups. Results of a second pairwise MANOVA established that more-skilled girls display less-complex play behavior during play with less-skilled playmates. When considered separately, none of the dependent variables were responsible for the variance. Rather, all three contributed Simultaneously to the significant overall multivariance. For boys, an overall multivariate analysis of variance was conducted but was not statistically significant. Boys do not alter their play during play with other four-year-old boys who display different levels of play complexity. Based on these findings, play actualizes the zone of proximal development for girls, not boys. Additional scholarship is needed in this area.