Archived soil samples from the beginning and end of three long‐term field trials conducted in central France were analysed for total uranium (U) and thorium (Th) contents to evaluate the effect of 15–30 yr of phosphorus (P) fertilizer treatments on the accumulation of these elements in the topsoil. For comparison, the soil samples were also analysed for total P. Three treatments were compared: no P application (P0), 26 kg P/ha/yr (P26) and 52 kg P/ha/yr (P52). Significant effects of P fertilizer were observed on U content and, to a lesser extent, on Th content as a result of the P52 treatment at two of the field trials. This effect was demonstrated both in the analyses at the end of the field trials [P52–P0: +0.25 and +0.44 mg U/kg soil, +0.58 (not significant) and +1.03 mg Th/kg soil] and when considering the changes in U and Th contents between the beginning and the end of the field trials (end–start: +0.18 and +034 mg U/kg soil, +0.35 and +0.45 mg Th/kg soil). The P fertilizer effect was also supported by the correlations of U and Th with total levels of P in the soil. However, in one of the three trials, no significant accumulation of U or Th because of fertilizer could be seen, suggesting either that less U and Th were applied using a different P fertilizer and/or that soil heterogeneity masked significant effects.