Inflaton Effective Potential from Fermions for General $\epsilon$

We accurately approximate the contribution of a Yukawa-coupled fermion to the inflaton effective potential for inflationary geometries with a general first slow roll parameter $\epsilon(t)$. For $\epsilon = 0$ our final result agrees with the famous computation of Candelas and Raine done long ago on de Sitter background, and both computations degenerate to the result of Coleman and Weinberg in the flat space limit. Our result contains a small part that depends nonlocally on the inflationary geometry. Even in the numerically larger local part, very little of the $\epsilon$ dependence takes the form of Ricci scalars. We discuss the implications of these corrections for inflation.