A description of rural school context variables for successful staff development and an empirical study of large-scale rural intervention are presented in this report. Drawing from research, literature, and practice based on nonrural populations, context variables are identified for implementing effective staff development intervention in rural schools. Document analysis, interviews, and observation were used to study 28 rural schools in 5 states--Arkansas, Louisiana, New Mexico, Oklahoma, and Texas. Stepwise regression was used to identify eight predictor variables for successful staff development, which include training time; socioeconomic profile of the student body; administrative participation; proximity to a higher education institution; provision of incentive; number of high school level teachers; number of administrators in the district; and presence of a trainer from a higher education institution. Three figures are included. (30 references) (LMI)