This paper presents a promising new item bias detection technique that derives practical appeal from its simplicity and economy, and provides data relating to the method's consistency across samples. The proposed regression on item characteristics (RIC) method is a straightforward and inexpensive approach to the study of item bias, based on transformed item difficulties and biserial correlations. It seems particularly well suited for situations of limited resources. It has certain methodological advantages over several competitors apt to be used in similar situations. The RIC method uses multiple regression to predict differences in group performance. Items identified by the proposed method as possibly biased are those items exhibiting group differences which cannot be accounted for by item difficulty, item discrimination, or overall group differences in ability. Results from this study suggest that the RIC method is a promising tool for item bias research although its consistency on small samples (N=125) has yet to be demonstrated. (Author)