Mounting evidence demonstrates that exclusionary discipline practices like suspensions and expulsions have long-term negative socio-emotional, behavioral, and academic consequences for the students who experience them, with evidence of spill-over effects for nonexcluded students. Restorative practice has emerged as a promising alternative to punitive discipline approaches, yet evidence is mixed on whether it can improve academic outcomes or curb racial disparities in school discipline. In a new conceptual model, we argue that the full potential of restorative practice can only be reached when it is (a) operationalized more directly within a socio-emotional framework; (b) responsive to more significant mental health needs; and (c) informed by a multifaceted understanding of how racism contributes to discipline disparities, both directly through interpersonal biases and indirectly through structural oppression. A revised conceptual model based on evidence from the broader literature, original data analyses, and pilot intervention results is advanced. [For the corresponding grantee submission, see ED620749.]