Manganese (Mn) exposure can result in parkinsonism. However, understanding of manganese neurotoxicity has been limited by the lack of a cumulative Mn biomarker. Therefore, the current goal was to develop Mn cumulative exposure indices (MnCEI), an established method to estimate cumulative exposure, and determine associations of MnCEI with blood Mn (BMn), fingernail Mn (FMn), and bone Mn (BnMn). We completed a cross-sectional study of 60 male Chinese workers. Self-reported occupational history was used to create two MnCEIs reflecting the previous 16 years (MnCEI16) and total work history (MnCEITOT). An in vivo neutron activation analysis system was used to quantify BnMn. BMn and FMn were measured using ICP-MS. Mean (standard deviation) MnCEITOT and MnCEI16 were 37.5 (22.0) and 25.0 (11.3), respectively. Median (interquartile range) BMn, FMn, and BnMn were 14.1 (4.0) μg/L, 13.5 (58.5) μg/g, and 2.6 (7.2) μg/g dry bone, respectively. MnCEI16 was significantly correlated with FMn (Spearman’s ρ = 0.44; p = 0.02), BnMn (ρ = 0.44; p < 0.01), and MnCEITOT (ρ = 0.44; p < 0.01). In adjusted regression models, MnCEI16 was significantly associated with BnMn (β = 0.03; 95% confidence interval = 0.001, 0.05); no other biomarkers were associated with MnCEI. This suggests BnMn may be a useful biomarker of the previous 16 years of Mn exposure, but larger studies are recommended.