Background: Providing preschoolers at-risk for reading difficulties with additional support is of increasing interest in early childhood education. However, the research on programming in preschool for this additional support is limited and yields inconclusive findings. Objective: The current studies explore different grouping configurations in a supplemental tutoring program for at-risk preschoolers in order to provide early childhood educators with guidance on grouping strategies for use in their supplemental instruction. Methods: Two grouping configurations are examined via two studies. In Study 1, 45 at-risk preschoolers (18 boys, 27 girls) were selected and randomly assigned to a one-on-one tutoring or paired tutoring condition. In Study 2, 54 at-risk children (31 boys, 23 girls) were selected and randomly assigned to one of two pairing conditions: with a highly-skilled peer or with a similarly low-skilled peer. In each study, children received tutoring that supplemented the classroom instruction twice a week over the academic year. Results: In Study 1, children in both conditions made similar gains on the alphabet knowledge and phonological awareness measures and the one-on-one group outperformed the paired group on receptive vocabulary but the effect size was small. In Study 2, the children in the matched-pairing condition evidenced a trend toward greater gains than those paired with high-skilled peer on the phonological awareness measure but not on alphabet knowledge and receptive vocabulary measures. Conclusion: The results of studies hold promise for achieving optimal outcomes by providing supplemental instruction to the maximum number of preschoolers using a dyad model instead of the typical one-on-one model.