A good test is one that has at least three qualities: reliability, or the precision with which a test measures what it is supposed to measure; validity, i.e., if the test really measures what it is supposed to measure, and practicality, or if the test, no matter how sound theoretically, is practicable in reality. These are the sine qua non for any test including tests of language proficiency. Over the past fifty years, language testing has witnessed three major measurement trends: Classical Test Theory (CTT), Generalizability Theory (G-Theory), and Item Response Theory (IRT). This paper will provide a very brief but valuable overview of these trends. It will then move onto a brief consideration of the most recent notion of Differential Item Functioning (DIF). It will finally conclude that the material discussed here is applicable not only to language tests but also to tests in other fields of science.