Background. High levels of fibroblast growth factor 23 (FGF23) are associated with mortality and progression of chronic kidney disease (CKD). Reducing dietary phosphorus intake lowers FGF23 secretion in healthly individuals, but there is little data on its effects in patients with pre-dialysis CKD.Methods. Using a 2 × 2 factorial design, we randomly assigned 16 normophosphataemic CKD stage 34 patients to receive a 2-week treatment with either lanthanum carbonate 1000 mg three times daily or placebo, and to ingest a tightly controlled diet containing 750 or 1500 mg of dietary phosphorus daily. We analysed serial measurements of FGF23, parathyroid hormone, serum phosphate and calcium, and 24-h urinary phosphate and calcium excretion using repeated-measures analyses.Results. Compared with the 1500-mg phosphorus diet, patients assigned to the 750-mg diet had greater reduction in 24-h urinary phosphate excretion (66% vs. 29%; P < 0.0001). Lanthanum-treated patients experienced a significant reduction in 24-h urinary phosphate excretion compared with baseline (64%; P < 0.0001), but the difference compared with placebo did not reach significance (64% vs. 31%). Despite the significant reductions in 24-h urinary phosphate excretion, no group demonstrated a significant reduction in FGF23 levels; FGF23 levels actually increased significantly in the 1500-mg diet plus placebo group, suggesting dietary phosphorus loading.Conclusions. Although dietary phosphorus restriction and lanthanum lowered urinary phosphate excretion consistent with a rapid decrease in phosphorus absorption, inducing a reduction in FGF23 levels in CKD patients may require interventions with a longer duration than in healthy volunteers. © The Author 2010. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.