The effect of alternating criterion-referenced tests (CRTs) with norm-referenced tests (NRTs) when evaluating schools was studied in a sample of 242 Louisiana public elementary schools accounting for over 18,000 third graders tested in 1989. The study used five separate multiple regression models, each producing studentized residuals used as school effectiveness indicators (SEIs). The SEIs were used to classify schools as effective, average, or ineffective. Each school was classified according to 10 different models. Cross-classification results were analyzed separately for: (1) CRT language arts and NRT language; (2) CRT language arts and NRT reading; and (3) CRT mathematics and NRT mathematics. Each comparison was tested with the kappa z-test. These tests were found to be significant beyond the 0.05 level. Magnitude measures were generally moderately consistent for CRT-NRT comparisons. Findings do not support alternating test modes when evaluating schools. Five tables present study data, and 19 references are included. (Author/SLD)