Background: Dexamethasone (D) improves acute nausea and vomiting when given pre chemotherapy (CTh), and continued D improves emesis after highly emetogenic CTh. There is lack of consensus about the optimal antiemetic regimen for delayed control after moderately emetogenic CTh. Many patients complain of symptoms while on D that may impact on QOL. Methods: A randomized double-blind cross-over trial compared D versus placebo (P) for delayed emesis in CTh-naïve women with breast cancer treated with moderately emetogenic CTh. All patients received IV granisetron and IV D pre-CTh and oral granisetron on day 2. Patients were randomized to oral D (4mg bid) or P for 48 hours, and changed to the alternative treatment for cycle 2. Primary endpoints were: (i) patient preference; (ii) difference between cycles in change in QOL (EORTC-QLQ-C30) from day 1 to 8. Secondary endpoints were emesis (Functional Living Index-Emesis; patient diaries) and the Dexamethasone Symptom Questionnaire. Results: Mean age of the 94 women recruited was 51yrs (range 27-76); 15 received AC and 79 FEC CTh. Thirteen women withdrew pre-cycle 2 with no difference in rates or reasons for withdrawal between arms. Thirty-one women preferred P and 37 preferred D (54% of those with a stated preference; 95% CI: 42-67%); 12 stated no preference. Of those preferring D, 54% rated it ‘much better’, compared with 39% for P. Patients had increased symptoms and decreased QOL from D1 to D8 in both cycles. There was greater decrease in global QOL (p=.06) and greater increase in pain (p=.04) when patients received D. No other symptom/QOL domains were significantly different. A non-significant decrease in vomiting intensity and delay in onset of vomiting was observed when receiving D, with no difference in nausea. Conclusions: No difference was found in patient preference, QOL or symptoms regardless of whether D or P was used for delayed emesis after moderately emetogenic CTh. We suggest continued D be omitted from initial antiemetic regimens for moderately emetogenic CTh, and offered selectively to patients with poor initial control of delayed nausea and vomiting.