Alkylresorcinols (AR) have been established as short/medium-term biomarkers for whole grain (WG) wheat and rye intake; and AR metabolites, 3,5-dihydroxybenzoic acid and 3-(3,5-dihydroxyphenyl)-propanoic acid, have been suggested as complementary biomarkers to AR. The present study examined the medium-term reproducibility and relative validity of urinary AR metabolites as biomarkers for WG and cereal fibre intake. A total of sixty-six free-living Swedes completed 3 d weighed food records and provided single 24 h urine collections and morning urine spot samples on two occasions, 2-3 months apart. The medium-term reproducibility of urinary AR metabolites was moderate when assessed in 24 h collections and lower in creatinine (CR)-adjusted morning urine. Mean AR metabolite 24 h excretions correlated well with total WG (r(s) 0·31-0·52, P < 0·05) and cereal fibre (r(s) 0·46-0·58, P < 0·001) intake on both occasions. As expected, correlations with WG (r(s) 0·28-0·38, P < 0·05) and cereal fibre (r(s) 0·35-0·42, P < 0·01) were weaker for mean CR-adjusted AR metabolite concentrations in spot samples of morning urine, although the adjusted concentrations correlated well with 24 h urinary excretion (r(s) 0·69-0·73, P < 0·001). Adjustment for intra-individual variations substantially improved the correlations between intake and excretion. These findings suggest that urinary AR metabolites can successfully reflect the medium-term intake of WG and cereal fibre when adjusted for intra-individual variation in this population, where rye was the major contributor to high WG intake. The performance of urinary AR metabolites as medium-term biomarkers appears to be comparable to that of fasting plasma AR concentration in this population.