Children with autism demonstrate considerable heterogeneity in their social skills and, therefore, their school intervention needs. No single intervention is expected to address the needs of all children with autism. In addition, not all evidence-based school interventions can be provided to all children with autism at all times. Thus, there is a need to understand how best to combine, sequence, and individualize social skills interventions to meet the heterogeneous needs of these children. "Adaptive interventions" (AIs) are prespecified sequences of decision rules used to guide schools in how best to combine, sequence, and individualize social skills interventions. However, there are currently no empirically derived AIs shown to improve social skills in schoolchildren with autism; moreover, there is a dearth of literature on the acceptability and feasibility of schoolwide, multilevel AIs that combine both environmental-level and individual-level interventions. The purpose of this study is to understand the acceptability and feasibility of four AIs in a SMART (sequential multiple-assignment randomized trial) implemented by educators and parents. The AIs include environmental (Remaking Recess, classroom supports) and individual interventions (parent assisted, peer mediated). Thirty-three elementary-age students with autism (male = 76%, Hispanic = 73%) were educated in 21 classrooms across seven schools by 25 teachers and 24 teaching assistants. Treatment expectations, acceptability, feasibility, and implementation data were collected over 18 weeks. Results indicated respondents were agreeable to treatment changes, but perceived feasibility was average and implementation was moderate. A number of lessons learned and proposed changes for scaling up are discussed.