Background: Nausea/vomiting (N/V) not related to anti-cancer treatment is common in patients with advanced cancer. The standard approach to management is to define a dominant cause, and treat with an antiemetic selected through pathophysiologic knowledge of emetic pathways. High rates of N/V control have been reported using both etiology-based guideline-driven antiemetic regimens and an empiric approach using single agents in uncontrolled studies. These different approaches had never been formally compared. Methods: This randomized, prospective, open label, dose-escalating study used readily available antiemetics in accordance with etiology-based guidelines or single agent therapy with haloperidol. Participants had a baseline average nausea score of ���3/10. Response was defined as a ��� 2/10 point reduction on a numerical rating scale of average nausea score with a final score < 3/10 at 72 h. Results: Nausea scores and distress from nausea improved over time in the majority of the 185 patients randomized. For those who completed each treatment day, a greater response rate was seen in the guideline arm than the single agent arm at 24 h (49% vs 32%; p = 0.02), but not at 48 or 72 h. Response rates at 72 h in the intention to treat analysis were 49 and 53% respectively, with no significant difference between arms (0.04; 95% CI: -0.11, 0.19; p = 0.59). Over 80% of all participants reported an improved global impression of change. There were few adverse events worse than baseline in either arm. Conclusion: An etiology-based, guideline-directed approach to antiemetic therapy may offer more rapid benefit, but is no better than single agent treatment with haloperidol at 72 h. Clinical trial registration: Australian New Zealand Clinical.