Objective:  Our study investigated changes of knee laxities in athletes and non-athletes females and relationship between knee laxity and sex-steroid at menstrual cycle phases. Methods:  Forty six healthy females, twenty four athletes and twenty two non-athletes not on hormone contraceptive pills, had no previous knee injuries and with regular menstrual cycles for 3 consecutive months, participated in the study. Medial and lateral knee laxities were determined by varus-valgus tests at follicular, ovulatory and luteal phases. Serum level of relaxin, estrogen, progesterone and testosterone were determined by ELISA and radioimmunoassay. Results:  Knee laxities in athletes and non-athletes at 0° and 20° flexion were the highest in luteal phase with non-athletes possess greater laxity than athletes. Positive correlation between progesterone and relaxin levels with knee laxities were observed. Meanwhile, the levels of both hormones were highest in the luteal phase. Conclusion:  Increased medial and lateral knee laxities in athletes and non-athletes associated with high serum progesterone and relaxin levels in luteal phase may contribute toward increased risk of non-contact knee injury. However, lower knee laxity in athletes than non-athletes suggest that exercise could be a protective factor.