BACKGROUND: Low protein intake (LPI) has been suggested as a treatment for chronic kidney disease (CKD). However, protein intake is essential for bone health. METHODS: We studied the database of the National Health and Nutrition Examination Survey, 2005-2010. Basic variables, metabolic diseases, and bone density of different femoral areas were stratified into four subgroups according to different protein intake (DPI) (that is, <0.8, 0.8-1.0, 1.0-1.2, and >1.2 g/kg/day). RESULTS: Significant differences were found among all lumbar area bone mineral density (BMD) and T-scores (p < 0.0001). There was an apparent trend between a decreasing BMD in the CKD groups with increasing DPI in all single lumbar spines (L1, L2, L3, and L4) and all L spines (L1-L4). Compared with DPI (0.8-1.0 g/day/kg), higher risks of osteoporosis were noticed in the subgroup of >1.2 g/day/kg over L2 (relative risk (RR)=1.326, 95% confidence interval (CI)=1.062-1.656), subgroup >1.2 g/day/kg over L3 (RR = 1.31, 95%CI = 1.057-1.622), subgroup <0.8 g/day/kg over L4 (RR = 1.276, 95%CI = 1.015-1.605), subgroup <0.8 g/day/kg over all L spines (RR = 11.275, 95%CI = 1.051-1.548), and subgroup >1.2 g/day/kg over all L spines (RR = 0.333, 95%CI = 1.098-1.618). However, a higher risk of osteoporosis was observed only in the non-CKD group. There was an apparent trend of higher DPI coexisting with lower BMD and T scores in patients with CKD. For osteoporosis (reference:0.8-1.0 g/day/kg), lower (<0.8 g/day/kg) or higher DPI (>1.2 g/day/kg) was associated with higher risks in the non-CKD group, but not in the CKD group. CONCLUSIONS: In the CKD group, LPI for renal protection was safe without threatening L spine bone density and without causing a higher risk of osteoporosis.