Background Deficiency of vitamin D is commonly associated with hip fracture and treatment with vitamin D reduces hip fracture rates. Consequently, the demand for assays to measure 25-hydroxyvitamin D (25-OHD) has increased. The Nichols Advantage chemiluminescence protein-binding assay (CLPBA) for 25-OHD is a first-generation automated immunoassay with decreased turnaround time, reduced manual handling and non-radioactive label.Methods We compared the CLPBA to the DiaSorin radioimmunoassay (RIA) and high-performance liquid chromatography (HPLC) for the measurement of 25-OHD using 161 samples from hip fracture patients and samples before and after institution of ergocalciferol (vitamin D-2) therapy.Results A negative bias for the CLPBA at concentrations below 30 nmol/L and a positive bias at 25-OHD values above 30 nmol/L compared with the RIA resulted in diagnostic discordance for one in three samples when using 30 and 50 nmol/L as decision limits. HPLC analysis confirmed the presence of a negative bias for the CLPBA at low values. Both immunoassays under-estimate 25-hydroxyvitamin D-2.Conclusions The discordance between 25-OHD values may be due to differences in standardization of each assay relative to HPLC. Our results emphasize the need for assay-specific clinical decision limits.