Self-extinction of timber after the removal of an external incident heat flux is necessary to ensure that structures with exposed wood do not contribute to a fire after the building contents are fully consumed. Two series of tests with solid wood and cross-laminated timber were conducted to study the conditions leading to self-extinction. The heat fluxes and heating duration were varied systematically. For each test, the evolution of the in-depth temperature distribution was measured using thermocouples while the mass loss rate was measured using a load cell. The experiments were allowed to reach a steady-state mass loss rate. The in-depth temperature gradient, which represents in-depth heat losses from the pyrolysis zone, was calculated from thermocouple measurements until it reached steady-state. As the external incident heat flux increases, the mass loss rate as well as the in-depth thermal gradient reach a steady-state value. Once the steady-state mass loss rate and temperature gradient were attained, the external heat flux was removed and the flame was allowed to extinguish. Over the entire range of heat fluxes (30-100kW/m2) and heating periods used in the tests, and for all the wood types used, self-extinction was observed.