[Objective] To analyze the correlation of serum creatinine (Cr), hemoglobin (HGB) and body mass index (BMI) with osteoporosis in middle-aged and elderly Tibetan residents in Xizang and its predictive value for osteoporosis. [Methods] Between November 2021 and August 2023, using multistage random cluster sampling method information on demographics, we collected blood indices and bone mineral density from 564 middle-aged and elderly Tibetan residents in Ali (Pulan, Zada, Ritu, and Rezhe counties), Nagchu (Nagchu, Nyima, Amdo, Shuanghu, and Bango counties), Shannan (Lunzi county), and Lhasa (Nimu county), Tibetan regions. They were divided into osteoporosis group (90 cases) and non-osteoporosis group (474 cases) according to whether they suffered from osteoporosis or not. Pearson correlation method was used to analyze the correlation between bone mineral density (BMD) and each variable. Binary Logistic regression was used to analyze the related influencing factors of osteoporosis in middle-aged and elderly Tibetan residents, and the regression equation model of predicting the risk factors of osteoporosis was constructed. The fitting degree of the model was tested by Hosmer-Lemeshow test model. Receiver operating characteristic (ROC) curve was used to evaluate the predictive efficacy of serum Cr, HGB and BMI combined tests on osteoporosis in middle-aged and elderly Tibetan residents in Xizang. [Results] The prevalence of osteoporosis was higher in females than in males. Among the osteoporosis group, height and weight, alanine aminotransferase, glutamine aminotransferase, urea nitrogen, uric acid, total cholesterol, triglyceride, low-density lipoprotein cholesterol, homocysteine, serum Cr, HGB and BMI were significantly lower (P<0.05). Correlation analysis showed that age was negatively correlated with BMD, while body weight, serum Cr, HGB and BMI were positively correlated with BMD (P<0.05). Multivariate binary Logistic results showed that age (OR=1.085), serum Cr (OR=0.891), BMI (OR=0.475), and HGB (OR=0.966) were the influencing factors for the occurrence of osteoporosis (P<0.05). To establish a prediction model for the risk of osteoporosis, Logit(P)=6.265+ 0.086×age + (-0.109×Cr)+ (-0.020×HGB)+ (-0.213×BMI). The area under the ROC curve was 0.932, the sensitivity was 88.9%, and the specificity was 82.5%. Hosmer-Lemeshow test model had a good fit (P=0.960). According to Logit(P) model, L score=-29.413+ (-0.404×age)+ 0.512×Cr+ 0.094×HGB+ BMI could be calculated. When L score >8.9, the patient could be diagnosed with the disease. [Conclusion] The decrease in serum Cr, HGB and BMI is a risk factor for osteoporosis in Tibetan aged residents. Age was negatively correlated with BMD, while serum Cr, HGB and BMI were positively correlated with BMD. The combined model of age, serum Cr, HGB and BMI can significantly improve the diagnostic efficiency of osteoporosis in the elderly Tibetans in Xizang.