Hemoglobin concentration differs by sex, possibly affecting any association between hemoglobin and frailty. This study aimed to evaluate the potential interaction effect of hemoglobin and sex on frailty in Chinese older inpatients. A cross-sectional study was conducted between February 2015 and November 2017 in a tertiary hospital. Frailty was defined by the Fried phenotype. Hemoglobin concentration was measured with a standard procedure. Covariates included demographics, clinical characteristics, and serum biomarkers. Logistic regression was applied to examine the association between hemoglobin concentration and frailty. The relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), and synergy index (SI) were used to evaluate the additive interaction. A total of 619 older inpatients [mean age 69.26±7.44 years; 334 men, 285 women] were included. The mean hemoglobin concentration was significantly lower in the elderly who were frail (11.9 g/L in frail versus 13.1g/L in non-frail; p<0.001). In the multivariable regression models, lower hemoglobin in patients was significantly associated with frailty (adjusted odds ratio (OR) = 2.51, 95% CI:1.37, 4.60). The stratified analyses indicated that lower hemoglobin was associated with frailty among older inpatients with different characteristics. Female inpatients with lower hemoglobin had the highest risk of frailty (adjusted OR=6.43, 95%: 2.38, 17.3); there were interactions between hemoglobin and sex on the development of frailty (RERI=4.30, 95% CI=-1.41, 10.01; AP=0.67, 95% CI=0.37, 0.97;SI=4.80, 95% CI=1.22, 18.84). Our study provided evidence that sex and lower hemoglobin have an interaction effect on frailty; it is suggested that clinicians may consider sex-specific strategies for the elderly to conform the concept of precision medicine. © 2021 Li et al.