Non-native earthworm invasions in north-temperate North America cause substantial adverse effects to hardwood forest ecosystems. Quantification of invasions is necessary for understanding impacts and identifying remnant earthworm-free areas, but existing sampling techniques are effort-intensive and/or environmentally damaging. We: (1) developed and applied a protocol that allows rapid classification of earthworm invasion into five stages based primarily on visual assessment of the forest floor, (2) sampled earthworms to test whether the protocol's stages can predict invasion by different species, and (3) assessed relationships between individual forest floor characteristics and presence of different earthworm species. Based on differences in biomass among points assigned to different stages, the 5-stage classification protocol accurately identified the onset of invasion by Lumbricus rubellus and Lumbricus terrestris, the species of greatest management concern in the northern Midwest. Except for middens as a predictor of L. terrestris presence, no forest floor variable was useful by itself for assessing invasions. The 5-stage protocol provides an efficient approach for assessing earthworm invasions in hardwood forests of the U.S. northern Midwest, can be implemented with minimal training, and serves as a blueprint for similar protocols in other regions experiencing earthworm invasions.