This study examines correlations between the Metropolitan Achievement Tests, seventh edition (MAT-7), and analogous portions of the 1995 Ohio Ninth-Grade Proficiency Tests. The MAT-7 scores of 156 eighth-grade students who had completed both test batteries were paired with complementary sections of the Proficiency test. Correlations between the tests were 0.52 for reading, 0.63 for mathematics, 0.25 for language/writing, and 0.58 for social studies and citizenship. All were considered significant. It is argued that even relatively high correlations between a test in question and an established test like the MAT-7 do not constitute validity. Validity comes only with the accumulation of results from different sources. Implications of comparing a high-stakes test like the Ohio Proficiency Test with a test administered in low-stakes circumstances are discussed. (Contains one table and nine references.) (Author/SLD)