The main purposes of this study were to examine first of all whether students' behaviour is affected by factors such as sex, academic achievement, personality type and social and economic status. Then, how the students' academic achievement influenced by the learning behaviour? The samples used in this work were taken from five hundred and sixty-three eighth grade students in four junior high schools respectively in the Taipei City. Gilford Personality Test, Investigations of Personal Basic Information and Junior-High-School-Students Learning Behaviour Inventory were conducted in this work. The obtained data were statistically analyzed by two-facts analysis of variance and stepwise multiple regression analysis. The results of the study were as follows: 1. Sex makes difference- -schoolgirls were always superior to schoolboys in the learning behavior. 2. The academic achievement played an important role, as the learning behavior was concern. No interaction existed between the academic achievement and the sex with the only exception of study skills. Generally speaking, students with high grade usually did better in learning behavior than those with middle grade did, while students with poor grade did the worst. 3. Personality maked significant difference in learning behavior. Students of calm type and directive type had better learning behavior than those of average type did, while students of black and eccentic type did the worst. 4. There was no significant difference between students of different SES in learning behavior, except that students with upper-class performed better than those with low-class of SES. 5. There was medium and significant relationship between academic achievement. Nine dimensions of learning behavior time schedure, study skills; examination-taking skills, attention, and voluntary study can account for and predict one third variance of students' academic achivement.