Weed suppression is one possible benefit of including cover crops in crop rotations. The late spring planting date of dry beans allows for more growth of cover crops in the spring. We assessed the influence of cover crops on weed dynamics in organic dry beans and weed seed persistence. Medium red clover, oilseed radish, and cereal rye were planted the year before dry beans; a no-cover-crop control was also included. After cover-crop incorporation, common lambsquarters, giant foxtail, and velvetleaf seeds were buried in the red clover, cereal rye, and no-cover control treatments and then retrieved 0, 1, 2, 4, 6, and 12 mo after cover-crop incorporation. Dry beans were planted in June and weed emergence and biomass measured. Eleven or more site-years of data were collected for each cover-crop treatment between 2011 and 2013, allowing for structural equation modeling (SEM), in addition to traditional analyses. Cereal rye residue increased giant foxtail and velvetleaf seed persistence by up to 12%; red clover decreased common lambsquarters seed persistence by 22% in 1 of 2 yr relative to the no-cover-crop control. Oilseed radish and incorporated cereal rye rarely reduced weed densities. When red clover biomass exceeded 5 Mg ha−1, soil inorganic N was often higher (5 of 6 site-years), as were weed density and biomass (5 and 4 of 12 main site sample times, respectively). Using SEM, we identified one causal relationship between cover-crop N content and weed biomass at the first flower stage (R1), as mediated through soil N at the time of dry bean planting and at the stage with two fully expanded trifoliates. Increasing cover-crop C : N ratios directly reduced weed biomass at R1, not mediated through changes in soil N. Cover crops that make a significant contribution to soil N may also stimulate weed emergence and growth.