Wildland fires are an extremely costly and deadly problem. Crown fires, where live foliage ignites and burns, are particularly unpredictable -in part because live fuel ignition and combustion is poorly understood. Many wildland fire models assume radiation is the controlling heat transfer mechanism. However, there is a growing indication that radiation is insufficient to ignite the small, thin fuel particles that carry a wildland fire and that convective heating and flame bathing is a critical component. Unfortunately, ignition by convection heating of any fuel is poorly understood. Ignition of live forest fuels by any means is also completely unknown due to complicated moisture content and fuel chemistry considerations. To gain some insight into the wildland fire problem, an apparatus was built using two 6.5 kW electrical heaters to heat gas (air, nitrogen, etc.) over a range of temperatures from ambient up to 1200°C. The flow rate of these "airtorches" is adjustable. This apparatus was used to convectively ignite a range of both live and dead forest fuels. Fuels from all over the United States where used including Southern California, Utah, Florida, and Montana. To examine ignition threshold conditions and to have distinguishable differences in ignition times, air temperatures of 500°C and 600°C were used. The airflow rate varied slightly from 1.3 m/s to 1.4 m/s due to the density difference. Because live forest fuels contain large amounts of water, the evolution of both water and carbon dioxide was measured with time using a differential gas analyzer. Flaming ignition was seen for all dead fuels at 500°C, but the live fuels mostly showed glowing ignition. At 600°C, all fuels showed flaming ignition within 1-26 sec. Interestingly, all live fuels were still actively releasing water at ignition, implying there are steep temperature gradients within these physically thin fuels (i.e. not thermally thin). Simple heat transfer analysis in conjunction with the water evolution information was used to help explain the differences in ignition times due to fuel geometry.