To compare 3 perioperative feeding regimens and their effect on anesthetic complications, manure output, and colic proportion in healthy horses. 45 horses presenting for elective orthopedic procedures were randomly assigned to 1 of 3 groups: not fasted (NF; continuous access to hay perioperatively), fasted muzzled (FM; 10-hour preoperative fast with slow refeeding postoperatively and muzzle placement), or fasted not muzzled (FNM; same as FM without muzzle placement). Anesthetic protocol was standardized. Outcomes compared between groups included anesthesia time, arterial oxygenation, duration of hypotension, perioperative manure output, time to first passage of manure postoperatively, pain scores, and colic proportion. Comparisons were made with a mixed model and Fisher exact test with statistical significance considered at P ≤ .05. No differences were seen in pain scores, oxygenation, hypotension, or colic between groups. Groups FM and FNM had a significantly greater mean reduction in postoperative manure weight (-81% and -70%; P = .003) and number of manure piles (-63% and -55%; P = .005) compared to group NF (-39% and -22%; P < .001; weight and piles, respectively). Mean ± SD minutes to passage of manure postoperatively was significantly shorter in group NF (238 ± 13 minutes) than groups FM (502 ± 174 minutes; P < .001) and FNM (444 ± 171 minutes; P = .003). Horses with continuous access to hay prior to and following recovery from anesthesia passed more manure and passed manure sooner after surgery than their fasted counterparts without detrimental effect on anesthetic parameters and postoperative complications. Continuous access to hay perioperatively supports manure production in healthy horses without increase in anesthetic complications.