The typical Geller-Seifter conflict paradigm for predicting clinical efficacy of anxiolytics is a mult VI/CRF schedule in which response rates in the CRF (conflict) portion are depressed by response-contingent electric shock. In 1-h sessions, anxiolytics raise the depressed conflict rates. Recently it was shown that replacing the single shock level with an arrangement whereby shock begins at zero and is increased with each response in the conflict portion produced more orderly data and facilitated training and maintenance of experimental subjects; chlordiazepoxide was the test drug. In the present study, those results are replicated in 30-min sessions, and the incremental paradigm is demonstrated to be as specific for anxiolytics as the standard Geller-Seifter paradigm. The possibility of very short sessions is suggested.