Background and aims Plants affect phosphorus (P) cycling through uptake and the mobilization of P from several soil pools into soil solution. The effects of seven cover crop species -three legumes (variable morphology), three cereals (variable domestication degree), one mustard (non-mycorrhizal) -on P cycling were compared in a greenhouse experiment. Methods Monocultures and legume-cereal mixtures were grown in an artificial plant growth substrate across three P input treatments (low P, manure, mineral fertilizer) to quantify changes in plant nutrients in aboveground and belowground biomass and properties of the plant growth substrate (pH, organic acids, enzyme activity, P). Results Legumes had the highest biomass, P uptake, and P mobilization potential (lower pH, higher organic acids and phosphatase activity) but cereals and mixtures mobilized more P than legumes. Biomass allocation to roots varied among species, with no trade-off between allocation to roots and P mobilization potential. Cereals had higher biomass, P uptake and N concentration in mixtures, whereas legumes had a mixed response in mixtures. Phosphorus concentration in the plant growth substrate affected plant growth and nutrient uptake but not P mobilization potential, with few differences between manure and mineral fertilizer. Conclusions Despite smaller effects on rhizosphere properties compared to legumes, cereals and mixtures had a greater impact on soil P and should affect P cycling more strongly when used as cover crops.