The importance of assessing language proficiency levels and the relationship of this priority to the teaching of Russian at the university level are discussed. Serious concerns about the value of language-specific proficiency guidelines are raised, and an argument is presented suggesting that language-specific guidelines may lead to undue emphasis on discrete-point testing. It is stressed that the importance of the proficiency testing program is not so much the test itself as the effect the program could and should have on the curriculum, and on what kind of information is being taught and how this information is taught. Reference is made to the poor performance of the most recent group of Russian language students evaluated by one tester to make the point that a firm foundation of language and structure should be laid before students become too fluent. Accuracy must be inculcated from the very beginning. The importance of teaching proper intonation is stressed, and attention is given to the Hypothetical Model of Relative Contributions (Higgs and Clifford) as an aid to curriculum design. (DJD)