Characteristic X rays of energies less than approximately 20 keV are of interest in radiobiology and radiation oncology. There is evidence that these low-energy photons produce higher relative biological effectiveness (RBE) and lower oxygen enhancement ratio (OER) relative to higher energies. Lower energy X rays also offer the advantage of healthy tissue sparing beyond the target treatment depth. Electronic brachytherapy systems that can deliver characteristic and bremsstrahlung X rays of varying energy are in clinical use as well as under development. We performed low-energy extrapolation ionization chamber dosimetry using two methods: 1. the exposure-to-dose method; and 2. the Burlin theory method combined with the extrapolation chamber method of Klevenhagen. We investigated fluorescent X rays emitted from seven metals: titanium (Ti, Z = 22); chromium (Cr, Z = 24); iron (Fe, Z = 26); cobalt (Co, Z = 27); copper (Cu, Z = 29); zinc (Zn, Z = 30); and molybdenum (Mo, Z = 42). X rays were produced by irradiation of the metals with a 55 kVp, 45 mA silver anode spectrum. The data obtained were air kerma rate (cGy/min), and radiation dose rate (cGy/min) in phosphate-buffered saline (PBS) solution and water. Air kerma rates ranged from 3.55 ± 0.10 to 14.36 ± 0.39 cGy/min. Dose rates ranged from 3.85 ± 0.10 to 16.96 ± 0.46 cGy/min in PBS and 3.59 ± 0.10 to 16.06 ± 0.43 cGy/min in water. Dose-rate energy dependence of both models was examined by taking a ratio of measured to Monte Carlo calculated dose rates. Dosimetry method 1 exhibited a linear relationship across all energies with a slope of 0.0127 keV⁻¹ and R² of 0.9276. Method 2 exhibited a linear relationship across all energies with a slope of 0.0467 keV⁻¹ and R² of 0.9933. Method 1 or 2 may be used as a relative dosimetry system to derive dose rates to water by using a second reference ion chamber with a NIST-traceable calibration for the molybdenum spectrum.