The rationale for seeding following high-severity wildfires is to enhance plant cover and reduce bare ground, thus decreasing the potential for soil erosion and non-native plant invasion. However, experimental tests of the effectiveness of seeding in meeting these objectives in forests are lacking. We conducted three experimental studies of the effectiveness of seeding with non-native and native species following three Arizona wildfires. Seeding treatments were largely ineffective in increasing vegetative cover or decreasing exposed bare ground. At one treatment at one fire, wheat seeding at the Warm Fire, senesced seeded annuals increased litter cover and resulted in lower bare ground values than unseeded controls. Only on one fire, the Warm Fire, did seeded non-native annuals establish well, resulting in 20-29% vegetative cover. On the other two fires, seeded cereal grains accounted for <3% cover. At all fires, native seeded species contributed between <1 and -12% vegetative cover. Vegetative cover on all treatments, including unseeded treatments, was at or near 40% the first year following fire, at all three study sites. Non-native species richness and abundance did not differ among treatments at any fire. This study adds to growing evidence that post-fire seeding is ineffective in enhancing post-fire plant cover and reducing invasive non-native plants.