Background <p>Clinical nurse preceptors play a central role in supporting nursing students and newly employed nurses during clinical practice. Teaching self-efficacy may be related to preceptors’ clinical teaching behaviors, yet little is known about how these psychological beliefs and observable behaviors co-occur within level-based profiles. This study aimed to identify latent profiles of teaching self-efficacy and clinical teaching behaviors among clinical nurse preceptors and examine factors associated with profile membership.</p> Methods <p>A cross-sectional survey was conducted from February to April 2026 at Guangyuan Central Hospital, China. Of 400 invited clinical nurse preceptors, 358 provided valid questionnaires, yielding an effective response rate of 89.5%. Participants completed a sociodemographic and occupational information questionnaire, the Chinese version of the Clinical Nurse Educator Skill Acquisition Assessment (CNESAA), and the Clinical Teaching Behavior Inventory (CTBI-23). Latent profile analysis was performed using the mean item scores of five teaching self-efficacy dimensions and six clinical teaching behavior dimensions as manifest indicators. Model selection considered fit indices, entropy, likelihood ratio tests, profile size, posterior classification probabilities, parsimony, and interpretability. Differences in participant characteristics across profiles were examined using χ² tests, and factors associated with assigned profile membership were analyzed using exploratory multinomial logistic regression.</p> Results <p>The mean item scores for teaching self-efficacy and clinical teaching behaviors were 3.70 ± 0.54 and 3.65 ± 0.53, respectively, indicating moderate levels slightly above the scale midpoint. A three-profile solution was retained as the more parsimonious and substantively interpretable model, although the four-profile solution showed better statistical fit and remained a plausible competing model. The retained solution comprised a low-level profile (<i>n</i> = 67, 18.7%), a moderate-level profile (<i>n</i> = 169, 47.2%), and a high-level profile (<i>n</i> = 122, 34.1%). Average posterior classification probabilities were high (0.978, 0.969, and 0.961 for the low-, moderate-, and high-level profiles after reordering), supporting classification accuracy. Compared with the low-level profile, older age, established employment, higher professional title, longer nursing and preceptorship experience, and more frequent participation in clinical teaching training were associated with membership in higher-level profiles.</p> Conclusions <p>Clinical nurse preceptors showed moderate levels of teaching self-efficacy and clinical teaching behaviors, slightly above the scale midpoint. The three profiles mainly represented ordered, level-based differences rather than qualitatively distinct patterns. These findings may inform stratified preceptor development, but should be interpreted cautiously because of the cross-sectional, single-center design and potential classification uncertainty.</p>