Spot fires caused by lofted embers (i.e., firebrands) can be a significant factor in the spread of wildland fires. Embers can be especially dangerous near the wildland urban interface (WUI) because of the potential for the fire to be spread near or among structures. Many studies have investigated the transport of lofted embers and the subsequent ignition of material on the ground, but knowledge regarding which fuel and environmental conditions control generation rates is sparse. Such information is needed to help inform ember transport models and to assess risks of ember generation for different fuel and environmental conditions. This work seeks to identify ember generation characteristics for different fuel characteristics and environmental conditions at multiple length scales. In laboratory experiments, dowels and natural samples of approximately 125 mm long were burned in a vertical wind tunnel. The species, moisture content, diameter, crossflow temperature, and crossflow velocity were varied. A factorial study with the time to generate an ember as the dependent variable found that diameter had the largest effect on the time required, followed by species. A subset of the data from the factorial study was used to compare manufactured dowels to natural samples. It was observed that natural samples of Douglas-fir took roughly 55% longer to generate an ember than corresponding manufactured dowel samples. At a larger scale, trees 2.1-4.7 m tall were burned in outdoor, semi-controlled conditions. Generated embers were collected in trays filled with water and on fire resistant fabric. The fire resistant fabric gives an indication of ember temperature upon deposition because only "hot" embers char the fabric. It was found that both the number of embers and char marks are significantly dependent on the fuel species. Of the species tested (Douglas-fir, grand fir, ponderosa pine, and western juniper), Douglas-fir generated the most embers per kilogram of mass loss during testing. Grand fir and western juniper generated the most char marks per kilogram of mass loss. It was observed that western juniper had the largest percentage of "hot" embers, with roughly 60% of the embers being hot enough to leave char marks. A technique was developed to identify important environmental factors that influence spotting distance at the forest-scale. This was accomplished by analyzing infrared images collected by the National Infrared Operations Program. Wind speed had the greatest influence on the propagation distance of embers. The comparisons of the observed maximum spotting distances with the predictions from Albini’s [1] model showed that modeled results were typically under predicted.