The COVID-19 self-test kit plays a pivotal role in public health by identifying active infections and enhancing individuals' access to testing. Its rapid results enable early detection of positive cases, contributing significantly to the control of disease transmission. This cross-sectional survey aimed to assess the knowledge, attitude, and behavior in using COVID-19 self-test kit and explored demographic characteristics, knowledge, and attitude as factors predicting COVID-19 self-test kit usage behavior among undergraduate students of the Faculty of Physical Education, Sports, and Health. The participants were selected by quota sampling technique. Data were collected from February 1 to 25, 2023. A total of 446 participants completed an anonymous online questionnaire voluntarily. Descriptive statistics, one-way analysis of variance and multiple stepwise linear regression were utilized for data analysis, using SPSS at a significance level of 0.05. The findings showed a high level of knowledge (71.3%) in using COVID-19 self-test kits, moderate scores in behavior (71.7%) and a neutral attitude (54.3%) towards their usage. Gender, year of study, department of study, and underlying disease exhibited noteworthy disparities in knowledge scores (all p<0.05). Similarly, gender, number of COVID-19 infections, and vaccination status were associated with attitudes (all p < 0.001). Furthermore, the department of study emerged as a significant factor influencing behavioral aspects (p = 0.013). Attitude (β = 0.427, 95% CI [0.455, 0.685], p < 0.001), being from the Department of Public Health (β = 0.199, 95% CI [2.388, 5.771], p < 0.001), knowledge (β = 0.168, 95% CI [0.316, 0.962], p < 0.001), being gender diverse (β = -0.089, 95% CI [-5.428, -0.157], p = 0.038), and being infected with COVID-19 twice (β = -0.083, 95% CI [-6.632, -0.015], p = 0.049) were significant predictors of COVID-19 self-test kit usage behavior. The identified predictors provide practical guidance for educational institutions and health authorities.