Background: A randomised study to assess the addition of apatorsen, an antisense oligonucleotide that inhibits Hsp27 expression, to docetaxel in patients with metastatic urothelial carcinoma (mUC) relapsed after prior platinum-based chemotherapy. Methods: Multicentre, phase II study with 1:1 randomisation to apatorsen (three loading doses at 600 mg intravenous followed by weekly doses) plus docetaxel (75 mg/m2 intravenous every 21 days) (A/D) or docetaxel alone. Overall survival (OS) was the primary end point with a P value <0.1 (one-sided) being positive. Progression-free survival (PFS), objective response rate (ORR), safety, and effect of Hsp27 levels on outcomes were secondary end points. Results: Patients randomised to A/D (n = 99) had improved OS compared to docetaxel alone (n = 101): HR: 0.80, 80% CI: 0.65-0.98, P = 0.0784, median 6.4 vs 5.9 months. PFS and ORR were similar in both arms. A/D had more incidence of sepsis and urinary tract infections. Patients with baseline Hsp27 levels <5.7 ng/mL had improved OS compared to those with levels ≥5.7 ng/mL. Patients with a decline or ≤20.5% increase in Hsp27 from baseline benefited more from A/D than those with >20.5% increase. Conclusions: A/D met its predefined OS end point in patients with platinum-refractory mUC in this phase II trial. This trial is hypothesis generating requiring further study before informing practice.