Mineral phosphate (P) fertilizers contain trace elements (TE) and can be a long-term source of these elements in soil. This study aimed to survey TE concentrations in mineral P fertilizers consumed in 25 of the EU-28 countries plus Norway (EU-28+1), to improve estimates of the EU wide input of TEs in agricultural soils. Different mineral P fertilizers (n = 414) were collected from EU-28+1 with a consumption-balanced sampling strategy. The samples were analyzed by ICP-MS for 21 elements (Na, Mg, Al, P, S, K, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Mo, Cd, Pb, Sb and U) that had adequate analytical quality control. Trace element concentrations in the P fertilizers were positively correlated with the P content for Cd, U, V, Sb, Cr, As and Ni. In addition, two groups of rock phosphates are likely the basis of P fertilizers in EU-28+1, i.e., with a high TE/P (n = 288) and low TE/P ratio (n = 126). The TE concentrations, relative to the aqua regia soluble concentration in EU arable soils, were highest for U, followed by Cd and then by other TEs. The Cd and U concentrations showed strong association but suggest that rock phosphates with two different Cd/U ratio are used. The Cd/P2O5 ratio varied significantly among countries and averaged 28 mg Cd kg-1 P2O5. An EU weighted mean is 30 mg Cd kg-1 P2O5, based on country average data and P consumption. Both means are below earlier EU estimates, likely because of sampling bias in those surveys. About 10% of the fertilizers exceed the current European Cd limit (60 mg Cd kg-1 P2O5). This survey illustrates the importance of a consumption-balanced sampling strategy to estimate TE inputs at a large scale and provides the data necessary for risk assessment of TEs in agricultural soils.