The reliability and validity of a naloxone rating scale (CINA) for assessing opioid dependence were studied in 75 opiate-abusing patients. Patients received 0.4 mg naloxone IV, and the naloxone-precipitated withdrawal effects were systematically measured at three five minute intervals using the 13-item CINA scale. The reliability coefficient alpha of the total withdrawal score was 0.81. The item scores and total score were correlated well except for nasal congestion, abdominal pain, and muscle pain. The inter-rater reliability of the total scores for the two raters was 0.93. The final methadone dose used in the treatment of individual patients was regressed on the CINA score. The simple regression equation was: METHADONE DOSE=11.2+ 0.72 CINA SCORE. This equation accounted for 69% of the variance of the final dose. By including a quadratic term, the prediction of the final methadone dose was improved by 3% (or a total of 72% of the variance); the regression equation was: METHADONE DOSE=4.36+1.26 CINA SCORE—0.01 CINA SCORE.2 The relatively large correlation (above 0.83) between CINA and the final methadone dose may be taken as an indication of the validity of the CINA score.