Abstract Background Assessment of free testosterone (FT) has been recommended in recent clinical practice guidelines as a biomarker for diagnosis and management of testosterone related disorders such as male testosterone deficiency due to hypogonadism and female androgen excess due to polycystic ovary syndrome (PCOS). FT is currently measured using direct commercial immunoassays, assays using equilibrium dialysis (ED) or other separation techniques. Though the ED approach is recognized as the recommended approach, it is laborious and technically challenging for routine clinical use. The CDC Clinical Standardization Program is developing a high throughput automated method using the ED procedure coupled with isotope dilution ultra-high-performance liquid chromatography tandem mass spectrometry (ID-UHPLC-MS/MS) that is suitable for routine patient care and large-scale epidemiologic studies. Methods Serum samples are dialyzed in a custom-designed multi-well plate against a protein-free HEPES buffer (pH 7.4) at 37 °C until equilibrium. After isolating endogenous FT from protein-bound testosterone by ED, isotope-labeled internal standard (13C3-testosterone) was added to the dialysate for quantification. Certified pure primary reference material (National Measurement Institute-M914) was used to prepare calibrators, enabling traceability and ensuring measurement trueness. FT was further isolated from the dialysate matrix using supported liquid extraction and a chromatographic separation from interfering compounds and quantitation by tandem MS. Results The measurement ranges covered 0.2–1000 ng/dL for testosterone, with the bias within ±5% and precision less than 10% CV. A total of 45 samples with a wide range of total testosterone (TT, 21–912 ng/dL) and SHBG (11–129 nmol/L) were analyzed and showed the suitability of this assay to measure the serum free testosterone levels in normal, hypogonadal males as well as in the majority of normal females and females with androgen excess. The application of 48-well format plate and automated liquid handler system significantly improves the throughput of sample preparation. In addition, comparison of the TT results measured with the presented TT method with calculated TT concentrations, suggest that on average calculation overestimates TT concentration by 34%. Conclusion The described high throughput method for FT allows for sufficiently accurate and precise measurement for routine applications including large epidemiologic studies to help with establishing reference intervals.