EVALUATION of the states of illness and change in patients is a major and continuing problem in psychiatry. Since ambiguities of definition and interpretation arise unavoidably when rating scales are relied upon, a number of investigators have attempted a shift to criteria that are more objective. For an altered focus, emphasizing interview form and process rather than the content of a patient's productions, the interaction chronograph has seemed a promising method. The method purports to predict a number of characterologic qualities by interpreting from a variety of measures, including the length of a subject's replies or silences, the number of times he interrupts the interviewer's questions, etc. Stable over time in both normal persons and in chronic, untreated patients, these measures and their correlates have been found by Chapple and others to vary with changes in clinical state, medication, and number of hours of psychotherapy.