This study examined whether FBA and function-based intervention methods developed in the U.S. would be effective if applied in a public school classroom in South Korea. Participants were three young children with severe mental retardation whose severe challenging behaviors put them at risk for removal from an inclusive kindergarten program. The study was conducted in two phases, and all conditions were conducted within the context of naturally occurring classroom activities. In Phase 1, FBAs identified that all three children behaved better when they had access to preferred activities and very poorly when they were engaged in non-preferred activities. In Phase 2, a function-based intervention in which instruction was embedded within preferred activities produced improvements in the behavior of all three children. Probes conducted at the children's residence showed good generality to a different setting. Finally, on a social validity questionnaire, teachers who implemented the intervention gave very positive ratings to the FBA and function-based intervention process. (Contains 1 table and 4 figures.)