Quality oral self‐care is a key element in maintaining oral health, which is important for students' general health, their overall well‐being and learning. The cluster randomised controlled trial tested the following hypotheses: (1) self‐determination theory (SDT)‐guided dental education is superior to conventional dental instruction in modifying oral self‐care in adolescents; (2) after the discontinued dental education, improved oral self‐care is only maintained for short rather than long term and (3) multiple predictors explain variations in adolescents' oral self‐care at different observation periods. The intervention group (N = 97) received three face‐to‐face educational sessions to facilitate adolescents' intrinsic motivation, while the control group (N = 99) had one conventional dental instruction session. Dental plaque scores (% of tooth area covered by dental plaque) indicated a lack of oral self‐care. Linear multivariable models tested the following predictors of oral self‐care at different observation periods: socio‐demographics (sex, socio‐economic status, school) and self‐determination attributes (autonomy, relatedness, competence). Results indicated that from baseline to the 6‐month follow‐up, dental plaque scores decreased (oral self‐care improved) in the intervention group but not in the control group. At the 12‐month follow‐up point, there were no significant differences in mean plaque scores between the study groups. Baseline plaque levels (β = 0.807), the type of dental instruction (theory guided vs. conventional) (β = 0.208), relatedness (β = 0.106) and competence (β = 0.102) were significant predictors that explained 67.6% of the variance in dental plaque scores at the 12‐month follow‐up point. The theory‐guided education was superior to conventional verbal instruction in improving adolescent oral self‐care; however, this improvement was only maintained short term. Variations in adolescent oral self‐care at short‐ and long‐term observation points were explained by baseline oral self‐care levels and two SDT components: relatedness and competence.