In the mental health study one important problem is how to investigate and ask some sensitive questions. The paper introduces the Randomized Response Technique (RRT). The technique can settle the problem.Main points of the technique are: (1) Ask the respondent two questions

one sensitive

another an inocuous question. (2) According to actual conditions the respondent choices of random the two questions with a Randomized Device

then answering "yes" or "no" to the selected question without revealing to the interviewer which question was actually answered. (3) On the Total probability formula. We can calculate the probability of answering of sensitive questions.The technique enables a respondent to provide truthful information on a sensitive or highly personal question and yet not reveal to the interviewer his status on the question

so it isuseful to apply in the mental health study