The aim of the study is to investigate the effect of mathematics self efficacy belief on mathematics teaching anxiety. The participants of the study consist of 42 mathematics teacher candidates in the Department of Mathematics Education, Faculty of Education. A 5-point Likert scale which was developed by Umay (2001) was conducted to measure the students' mathematics self-efficacy beliefs, and a 5-point Likert scale which was developed by Peker (2006) was conducted to measure the students' mathematics teaching anxiety. Pearson correlation was used for the correlation between teacher candidates' average points from the both scales. The results revealed that there was a significant correlation mathematics self efficacy beliefs and mathematics teaching anxiety (p=. 01, r= -0.59). In this situation, it can be said that there is a negative and meaningful relationship between mathematics self efficacy belief and mathematics teaching anxiety. This result means that the students' mathematics teaching anxiety decrease as their mathematics self efficacy beliefs increase. On the other hand, it was seen that mathematics self efficacy belief explains 35% of the variance in teaching mathematics anxiety. On the other hand, it was seen that mathematics teacher candidates' mathematics teaching anxiety was low (approximately average 2 point), and mathematics efficacy beliefs is high (approximately average 4 point). When looked at the results of the current study, it seems that in order to decrease mathematics teacher candidates' mathematics teaching anxiety, their mathematics self efficacy beliefs must be increased. Mathematics self efficacy refers to the belief or perception that one is capable of organizing and executing the actions necessary to succeed at a given mathematical task. For this aim, their knowledge and abilities relating mathematical concepts and also their confidences about their mathematical abilities must be enhanced.