Early students of political tolerance projected a rapid rise in levels of tolerance, but subsequent research has failed to offer conclusive evidence regarding whether tolerance has, in fact, increased. The General Social Survey (GSS) included the same 15 dichotomous tolerance items in the period 1976–1998, seemingly permitting assessment of trends in tolerance through examination of a standard 0–15 scale. Unfortunately, the validity of these data is uncertain because we cannot rule out the possibility that changes in affect toward the five GSS target groups, not true changes in tolerance, drive movement in the longitudinal series. To address this problem, we reexamine the GSS data from the perspective of past discussions of the meaning of tolerance and intolerance. We argue that that the GSS scale captures two aspects of tolerance: whether respondents are tolerant or intolerant; and, among the intolerant, the breadth and depth of their intolerance. We further argue that the first of these dimensions can be measured validly using the GSS data. Our analyses reveal that, at most, tolerance has increased only marginally in the period 1976–1998.