The validity of various procedures for rank ordering school districts throughout a state on the basis of statewide test results was studied. Data were drawn from the 1987 administration of the Kentucky Essential Skills Test (KEST), which is given each spring to all children enrolled in public school grades kindergarten through 12 in 178 school districts in the Commonwealth of Kentucky. With the data, school districts were ranked, first using different aggregates of achievement scores and then using achievement scores adjusted for different combinations of demographic factors. The demographic factors and school-based indices included: percent economically deprived, percent local revenue, assessed property value, cost for instruction, current expenses, percent local education agency support, local financial index, teacher/student ratio, and attendance and dropout rates. Achievement variables included: total battery normal curve equivalents score (TBNCES), grade 5 TBNCES, grade 10 TBNCES, reading comprehension NCES, and mathematics score. These rankings were compared, and differences were noted. The various ranking procedures yielded widely varied ranks for most of the districts, regardless of whether non-adjusted or adjusted scores were used. Thus, any ranking procedure is likely to create inaccurate public and governmental perceptions of variation in educational programs, and may lead to misgiven conclusions regarding the programs' quality. Seven data tables are provided. (Author/TJH)