INTRODUCTION: Dyspnea is frequently encountered in end-stage diseases even when reversible issues are addressed. Which clinical factors best define patient subpopulations that will most predictably benefit from opioids in this clinical setting? METHODS: Thirty-eight patients with refractory dyspnea were randomized to an 8-day crossover trial of 4 days of 20 mg sustained release morphine or placebo, switching arms on day 5 (Clinical Trial Registry Number: ACTRN012607000075482). Dyspnea was measured on a 100-mm visual analogue scale (VAS). Day 4 and day 8 morning and evening VAS scores were the primary outcome. Correlation between baseline dyspnea and response to opioids was explored; potentially important clinical predictors tested with two-sided Student's t test. RESULTS: In this exploratory study, no relationship could be defined between baseline dyspnea and response to opioids (Spearman correlation 0.03, p = 0.88). The study was not powered to define other predictors, but younger age, better functional status, and significant cardiac findings on entry to the study deserve further prospective evaluation in a larger cohort. DISCUSSION: Phase 4 pharmaco-vigilance trials are needed in palliative care to define people who are most likely to experience a net benefit from treatment such as opioids for refractory dyspnea.