Explaining restoration failure can be as important as touting success. We used a series of studies to understand the failure of techniques commonly used to restore wetland prairies in the Willamette Valley of western Oregon. Burning, fallowing, and solarization (covering tilled plots with plastic sheeting to heat the soil) had pronounced first-year effects on several individual species, but either did not reduce overall pest plant abundance or reduced the abundance of native species as well. The 34% overall plant cover in solarized plots was the only significant difference from the 60% cover present in control plots. All first-year responses essentially disappeared by the second year. These measures had little lasting effect on pest and other exotic plants because many survived treatment and resprouted. In addition, treatments had little effect on the number of seeds in the soil, leaving a pool of immediate and potential regeneration. Specific control measures of target plants, such as hand removal and repeat maintenance after initial treatments, should prove more successful. In a second study, three mixtures of native species sown into fallowing treatment plots had low emergence rates of 1% -7%, despite high seed viability, and produced only 0% – 3% cover. Native species should be selected and sown at densities high enough to lead to significant numbers of surviving seedlings, especially in the face of competition from surviving pest plants.