A systems trial was designed to evaluate forage characteristics within mixed-species pastures consisting of (i) endophyte-infected tall fescue (Festuca arundinacea Schreb.; E1) mixed with common bermudagrass [Cynodon dactylon (L.) Pers.] and other forages; (ii) endophyte-free tall fescue (E2) overseeded into dormant common bermudagrass; and (iii) orchardgrass (OG; Dactylis glomerata L.) established under the same conditions as E2. The E2 and OG pastures were grazed with either twice weekly (2W) or twice monthly (2M) rotation schedules, but E1 was grazed only as 2M. Across 41 sampling dates (2000 through 2003) the mean forage mass across all forage systems was 3809 kg ha 21 , and there was an interaction of forage system and sampling date (P 5 0.001). In vitro dry matter disappearance (IVDMD) and crude protein (CP) varied (P , 0.0001) with sampling date in seasonal patterns that were generally predictable. Frequencies of tall fescue in E2 and E1 pastures increased (P , 0.10) over years, reaching numerical maxima (61 to 72%) at the end of the trial. For OG, frequencies reached numerical maxima of 52 and 42% in 2W and 2M pastures, respectively, but then declined (P , 0.10) over time, ending at 39 and 24%, respectively. At the end of the trial, reinfection of OG pastures by rogue E1 plants reached a numerical maximum frequency of only 10%, and concentrations of total ergot alkaloids in tall fescue plants from E2 pastures were only about 25% of those for E1 pastures, thereby suggesting that pasture toxicity can be reduced substantially for at least 5 yr using these alternative forage systems.