Objectives: To establish multiple strains of Dichelobacter nodosus in two flocks of sheep and to assess the virulence of five of these strains. Methods: In experiment 1, sheep were challenged with five D. nodosus strains, which varied in both virulence and serotype. In experiment 2, consisting of four replicates (paddock groups), sheep were challenged with seven different strains of D. nodosus. In experiment 3, sheep were challenged with one of five D. nodosus strains. Results: In experiment 1, at 28 days post challenge, four of the five challenge strains were present. Multiple-strain infections were present in 27 feet, with isolates from three serogroups being recovered from 5 feet, and four serogroups from 1 foot. Challenged hind feet were more frequently affected (P < 0.001). In experiment 2, four of the seven strains were recovered from one replicate and three strains from the remaining three replicates. Significantly more hind feet were affected (262/471, 55.6%) than front feet (198/481, 42%) (P < 0.001). Clinically, in both experiments 1 and 2 the footrot resembled an intermediate form, despite the inclusion of a virulent strain of D. nodosus. In experiment 3, this virulent strain caused a higher prevalence of more severe footrot, a greater mean total foot score and, in Merino sheep, resulted in significantly lower weight gains (P < 0.05). Interaction between D. nodosus strain and breed occurred, with Polwarth sheep being significantly more severely affected by one strain than Merino sheep. Discussion: The clinical expression of multiple-strain infections has implications for both research and control of footrot. A novel method of control is proposed.