Abstract: Objective: This prospective cohort study was conducted in a healthy population to investigate the association between different obesity metabolic phenotypes and the risk of breast cancer in females. Methods: A prospective cohort study design was employed, enrolling female employees and retirees who participated in annual team physical examinations at Haikou Fourth People's Hospital from January 2015 to December 2024 (n = 11,564). A dynamic follow-up queue was constructed, and breast cancer incidence outcomes were monitored annually until January 2025. Baseline data on study participants were collected using questionnaire surveys, body measurements, traditional Chinese medicine (TCM) constitution assessments, and laboratory tests. Participants were assigned into four phenotype groups based on their metabolic status and body mass index (BMI) levels: metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUNW), metabolically healthy obesity (MHO), and metabolically unhealthy obesity (MUO). Multiple correspondence analysis was used to explore the correlation between TCM constitution types and metabolic obesity phenotypes. The Cox proportional hazards regression model was used to assess the impact of different metabolic obesity phenotypes on the incidence of breast cancer in females. Results: The cumulative follow-up of 11,564 subjects was 109,703.08 person-years, with a median follow-up time of 9.50 years. A total of 175 new cases of breast cancer were identified, corresponding to an overall incidence of 159.52/10⁵ person-years. The incidence densities in the MHNW, MUNW, MHO, and MUO groups were 125.47/10⁵, 222.42/10⁵, 175.65/10⁵, and 242.95/10⁵ person-years, respectively, with statistically significant differences (χ² = 12.222, P = 0.007). The cumulative incidence rates for the four phenotypic groups were 1.19%, 2.10%, 1.68%, and 2.27%. The Log-rank test revealed a statistically significant difference between the groups (χ² = 12.298, P = 0.006). According to the multiple correspondence analysis, MHNW was associated with peaceful and phlegm dampness constitution; MUNW was associated with qi-stagnation constitution; MHO was associated with Yang-deficiency, damp-heat, and characteristic constitutions; and MUO was associated with qi-deficiency and blood-stasis constitutions. Logistic regression analysis of the relationship between TCM constitutions and breast cancer revealed that female participants with qi-deficiency, yang-deficiency, damp-heat, and qi-stagnation constitutions had a higher risk of breast cancer than those with a peaceful constitution (odds ratio = 1.208, 1.263, 1.155, and 1.565, respectively; P < 0.05). Multivariate Cox proportional hazards regression analysis of different metabolic obesity phenotypes and breast cancer risk showed that, after adjusting for the influence of TCM constitutions and other confounding factors, the incidence risk of breast cancer was higher in the MUNW, MHO, and MUO groups compared with the MHNW group (aRR₂ = 1.779, 1.330, and 2.577, respectively; P < 0.05). The MUO group had the highest risk, followed by the MUNW group, with the MHO group having the lowest risk. Conclusions: Our findings suggest that the risk of breast cancer is the highest in individuals with metabolically abnormal obesity, followed by those with metabolically abnormal normal weight, and is greater than that in individuals with MHO. TCM constitution further amplifies this risk, with specific constitutions closely associated with metabolic phenotypes.