A new method has been developed to impose different redox conditions in high-temperature-pressure experiments in cold-seal pressure vessels, at 800 °C and 2000 bars. Experiments were conducted by loading a metallic filler rod into the autoclave together with H2 sensor capsules, and pressuring the autoclave with H2O. Rod materials tested successfully were Co, Ti, and C (graphite). The oxidation of these rods produces H2, but because of diffusive H2 loss through the walls of the autoclave, the system may not be buffered with respect to H2. However, fH₂ quickly reaches a steady state value, and because fH₂ is easily measured by the hydrogen sensor method, the effect of the filler rods on the intrinsic fO₂ of the autoclave can be quantified. In order to produce oxidized conditions, Ar was used as the pressure medium and metal oxides, contained in Al2O3 tubes, were employed. By using either Ar or H2O as a pressure medium, a log fO₂ range of NNO -3.9 to NNO +4.6 can be imposed by this method, where NNO is the log fO₂ value of the Ni-NiO buffer. The ability to conduct long-run-duration experiments at high temperature and high fH₂ conditions is not possible with the traditional double-capsule technique because the buffer assemblage is consumed too quickly. However, run durations of up to 4 weeks with constant fH₂ at reduced conditions have been conducted using the filler-rod technique. This technique has been shown to be an effective method in controlling redox conditions in cold-seal autoclaves, and thus can be applied to investigating redox-dependent reactions in a wide range of geochemical systems.