This paper examines two questions. First, how stable is religiosity over time? Second, how does religiosity affect the stability of attitudes over time? I begin by discussing several reasons why religiosity might help to stabilize attitudes. Then, drawing on the 1992-94-96 National Election Study panel, I examine the stability of religious tradition, religious movement identification, church attendance, view of scripture, and the overall importance of religion. For the most part, these indicators are found to be fairly stable, though not uniformly so. I then examine the effect of religiosity on the stability of attitudes towards abortion. Religiosity is found to have no significant impact. I conclude by speculating on reasons why.