Nausea and vomiting remain important clinical problems occuring in 25 to 50% of patients receiving chemotherapy for cancer. Clinical trials comparing a new antiemetic drug, ondansetron, to metoclopramide have suggested improved control of nausea and vomiting but studies disagree on the magnitude of the treatment effect and its statistical significance. We combined evidence from randomized controlled trials in a meta-analysis of the efficacy and safety of ondansetron compared to metoclopramide in the prevention of acute (less-than-or-equal-to 24 hours) nausea and emesis associated with chemotherapy. Literature search identified six randomised controlled trials of ondansetron versus metoclopramide in an adult population. Study outcomes were the observed incidence of emesis (vomiting or retching) and patient-reponed grades of nausea after chemotherapy. For meta-analysis of each outcome we defined therapeutic success as complete protection (ie. zero episodes during 24 hours following chemotherapy). The relative odds of success (ondansetron/metoclopramide) was calculated for each trial and all trials combined. Results were expressed as a relative risk (RR) for zero emesis or nausea at 24 hours. The six trials reported on 705 patients (median age range 53-59 years; 57% female). Relative odds for complete control of emesis was greater than one in all trials but was nonsignificant (p>0.05) in two trials, including the largest trial. When trials were combined, summary odds ratios for control of emesis and nausea were greater than one (p<0.05). RR of zero emesis with ondansetron was 1.72 (95% CI 1.45 to 1.97) and was similar for nausea (RR= 1.78, 95% CI 1.39 to 2.13). In trials using high-dose cisplatin chemotherapy, higher rates of extrapyramidal affects and diarrhea were associated with metoclopramide (p<0.05) while headache was frequently associated with ondansetron (p<0.05). Combined clinical trial evidence supports the conclusion. that, relative to metoclopramide, ondansetron places patients at a much lower risk of nausea and emesis following chemotherapy with moderately or highly emetogenic regimens.