Field studies were carried out in Laingsburg, MI, from 2002 to 2004 on Houghton muck soil to assess the impacts of cover crops and soil fertility regimes on weed populations and celery yield. The cover crops were oilseed radish, cereal rye, hairy vetch, and a bare ground control. The fertility rates were full (180, 90, and 450 kg ha−1 nitrogen [N], phosphorus pentoxide [P2O5], and potassium oxide [K2O], respectively), half (90, 45, and 225 kg ha−1 N, P2O5, and K2O, respectively), and low (90 kg ha−1 N). Each cover crop treatment was combined with the low or half rate of fertilizer. An additional treatment with bare ground plus the full rate of fertilizer was added as standard practice. Treatments were maintained in the same location for the duration of the study. Major weed species were common chickweed, prostrate pigweed, shepherd's-purse, common purslane, and yellow nutsedge. Each year, oilseed radish consistently produced the greatest biomass and provided over 98% early season weed biomass suppression. Hairy vetch and cereal rye provided about 70% weed suppression in early spring. Soil fertility level affected weed populations during the 2004 growing season. In 2004, weed biomass in treatments without cover crops or with vetch increased when greater amounts of fertilizer were applied. Within individual fertility levels, higher celery yields were recorded in the oilseed radish plots. For example, in the low fertility rate, celery yield was 34.8, 29.2, 23.9, and 24.4 ton ha−1 in the oilseed radish, cereal rye, hairy vetch, and control plots, respectively in 2003. Overall, the results of this experiment indicate that when included in a system where hoeing and hand-weeding are the only weed control methods, cover crops can successfully improve weed management and celery yield on muck soils, allowing reduced fertilizer inputs.