Objective To explore the latent classes and influencing factors of health behavior decision-making in stroke patients, thereby providing evidence for formulating personalized intervention measures. Methods A convenience sampling approach was adopted to enroll 203 stroke patients. Then they were investigated by using a general information questionnaire, the Stroke Patient Behavior Decision-making Assessment scale, the Stroke Self-efficacy Questionnaire (SSEQ), the Stroke Recurrence Risk Perception Assessment subscale (SRRPA), and the Social Support Rating Scale (SSRS). Latent profile analysis was conducted to identify latent classes of health behavior decision-making, then univariate analysis and multivariate logistic regression analysis were performed to determine the influencing factors of latent profiles. Results The score of health behavior decision-making in stroke patients was (105.90±20.11) points, which could be divided into two latent groups:passive hesitant decision-making group (67.98%) and active decision-making group (32.02%). Multivariate logistic regression analysis showed that, age, number of comorbidities, score of SRRPA, SSEQ and SSRS were influencing factors of latent classes of health behavior decision-making in stroke patients (all P<0.05). Conclusion There is heterogeneity in health behavior decision-making among stroke patients, which can be divided into two latent classes. Healthcare providers can implement individualized interventions according to the influencing factors of different latent classes, so as to improve the health behavior decision-making level and clinical outcomes of stroke patients.