Technical Abstract: Quantifying long term effects of soil erosion on plant production and soil quality factors can aid in restoring degraded soils to production. The objectives of this study were to determine the long-term effects of topsoil removal (artificial erosion) on spring wheat (Triticum aestivum L.) yield, grain protein, and selected soil productivity factors. In 1982, the top 0, 6, 12, and 18 cm of topsoil was mechanically removed from a Williams loam (fine-loamy, mixed, superactive, frigid Typic Argiustoll). Separation berms between plots were planted to perennial grass. Plots from 1982 to 1999 were planted to spring wheat or fallowed. In 2000 and 2001, a wheat-pea (Pisum sativum L.) rotation was implemented. Soil samples were collected in 1998 to 36 cm in 6 cm increments to determine soil organic matter (SOM), particulate organic matter (POM), water stable soil aggregates (WSA), microbial biomass, mineralizable N, profile water content at spring planting, and pH. Wheat yields and grain protein were measured from 1998 to 2001. In 1998, there was no difference in SOM, POM, mineralizable N, or WSA in the top 12 cm of soil due to topsoil removal. However, there was an average loss of 5.3 g SOM kg-1 soil from the top 12 cm of all soil removal treatments since 1982. There was an increase in SOM and POM in the top 6 cm of soil where grass was planted in strips between plots, but no increase at the 6 to 12 cm depth. The removal of 6 cm or more topsoil reduced fungal biomass in the top 6 cm of soil; whereas, bacterial biomass was not reduced until 12 cm or more was removed. Topsoil removal had no significant effect on soil water at planting, nor did it affect wheat yields. Grain protein was reduced only 1 of 4 years. This plot site is nearly level with little or no runoff ensuring that all treatments had equal available water. Consequently, the use of commercial fertilizer has maintained crop yields, but SOM has continued to decline indicating that, under the best conditions, restoration of damage caused by erosion is unlikely with a conventionally tilled wheat-fallow rotation in semiarid regions.