The first part of the lecture presents three classical aspects of behavior therapy: therapeutic efficiency, utilisation of well defined methods, rapidity of treatment. Emphasis is brought to the importance of behavior analysis and to the dangers arisen from insufficient knowledge or ignorance of learning theories and fundamental principles of behavior analysis. One insists on experimental analysis which is the basis for methods of action having as consequence for the therapeut: controlling strategy of treatment and direct responsibility concerning success and failure. Then, actual limitations of behaviour therapy are described. The second part deals with one of the most important working hypothesis on human neurosis originating from laboratory and clinical research. Behavior therapy refuses to elaborate hypothetical deductive constructions, unless working hypotheses to be verified. Then the four essential clues to behavioural psychotherapy are formulated. The main methods of action presently utilized are presented: aversive methods, operational and systematic desensitization technics through reciprocal inhibition. Finally, some of the main criticism usually made on behaviour therapy are being discussed.