This paper was presented as part of a symposium on "Current Views of Reliability" at the 1960 meeting of the American Psychological Association Meeting. The author reviews then current trends in thinking about validity coefficients and reliability coefficients as having fundamental places in mental test theory. He notes his belief that the unreliability of test scores is a basic and characteristic feature of test theory. However, it is not the reliability coefficient that is basic; it is the relation between the examinee's fallible observed score and the unknown true score. Mental test theory is concerned with the scatterplot between observed score and true score for a group of examinees.