Background: Testosterone is an androgenic hormone that plays important roles in both males and females. The circulating levels of total testosterone vary from 1 to 1480 ng/dL. High-throughput immunoassays often lack accuracy in lower concentration ranges (below 100 ng/dL), particularly when used for females or children. To address this limitation, we developed a total testosterone LC-MS/MS assay on three instruments. Methods: Sample preparation began with the dilution and conditioning of 200 µL of serum. A supported liquid extraction cartridge was used to extract the analyte from biological matrices. Chromatographic separation was achieved using a C18 column with a runtime of 5 min per sample. This assay was validated on a Triple Quad 6500 and an API 4500 instrument. Results: Method validation was carried out according to the CLSI C62-ED2 guideline and our hospital protocol. The within-day coefficient of variation (CV) was less than 10% and the between-day CV was less than 15%. The assay had a limit of quantitation of 0.5 ng/dL with an analyte measure range of 2–1200 ng/dL. A comparison using Deming regression and Bland–Altman plots showed that this assay correlated well with a reference method. The results from the API 4500 and an Orbitrap were consistent with those from the TQ 6500. Both serum-separator tubes (BD) and serum-activator tubes were found to be suitable. Conclusions: We successfully developed and validated a robust total testosterone LC-MS/MS assay for routine clinical testing. This assay was harmonized across two triple quadrupole instruments and one high-resolution mass spectrometer.