Androgen-deprivation therapy in patients with prostate cancer who have relapsed with rising prostate-specific antigen concentration only (PSA-only relapse), or with non-curable but asymptomatic disease at diagnosis, could adversely affect quality of life at a time when the disease itself does not. We aimed to compare the effect of immediate versus delayed androgen-deprivation therapy on health-related quality of life over 5 years in men enrolled in the TOAD (Timing of Androgen Deprivation) trial. This randomised, multicentre, open-label, phase 3 trial done in 29 public and private cancer centres across Australia, New Zealand, and Canada compared immediate with delayed androgen-deprivation therapy in men with PSA-only relapse after definitive treatment, or de-novo non-curable disease. Patients were randomly assigned (1:1) with a database-embedded, dynamically balanced algorithm to immediate androgen-deprivation therapy (immediate therapy group) or to delayed androgen-deprivation therapy (delayed therapy group). Any type of androgen-deprivation therapy was permitted, as were intermittent or continuous schedules. The European Organisation for Research and Treatment of Cancer (EORTC) quality-of-life questionnaires QLQ-C30 and PR25 were completed before randomisation, every 6 months for 2 years, and annually for a further 3 years. The primary outcome of the trial, reported previously, was overall survival, with global health-related quality of life at 2 years as a secondary endpoint. Here we report prespecified secondary objectives of the quality-of-life endpoint. Analysis was by intention to treat. Statistical significance was set at p=0·0036. The trial was registered with the Australian New Zealand Clinical Trials Registry, number ACTRN12606000301561, and ClinicalTrials.gov, number NCT00110162. Between Sept 3, 2004, and July 13, 2012, 293 men were recruited and randomly assigned; 151 to the delayed therapy group and 142 to the immediate therapy group. There was no difference between the two groups in global health-related quality of life over 2 years from randomisation. There were no statistically significant differences in global quality of life, physical functioning, role functioning, or emotional functioning, fatigue, dyspnoea, insomnia, or feeling less masculine over the entire 5 years after randomisation. Sexual activity was lower in the immediate therapy group than in the delayed group at 6 and 12 months (at 6 months mean score 29·20 [95% CI 24·59-33·80] in the delayed group vs 10·40 [6·87-13·93] in the immediate group, difference 18·80 [95% CI 13·00-24·59], p<0·0001; at 12 months 28·63 [24·07-33·18] vs 13·76 [9·94-17·59], 14·86 [8·95-20·78], p<0·0001), with the differences exceeding the clinically significant threshold of 10 points until beyond 2 years. The immediate therapy group also had more hormone-treatment-related symptoms at 6 and 12 months (at 6 months mean score 8·48 [95% CI 6·89-10·07] in the delayed group vs 15·97 [13·92-18·02] in the immediate group, difference -7·49 [-10·06 to -4·93], p<0·0001; at 12 months 9·32 [7·59-11·05] vs 17·07 [14·75-19·39], -7·75 [-10·62 to -4·89], p<0·0001), but with differences below the threshold of clinical significance. For the individual symptoms, hot flushes were clinically significantly higher in the immediate group (adjusted proportion 0·31 for delayed therapy vs 0·55 for immediate therapy, adjusted odds ratio 2·87 [1·96-4·21], p<0·0001) over the 5-year period, as were nipple or breast symptoms (0·06 vs 0·14, 2·64 [1·61-4·34], p=0·00013). Immediate use of androgen-deprivation therapy was associated with early detriments in specific hormone-treatment-related symptoms, but with no other demonstrable effect on overall functioning or health-related quality of life. This evidence can be used to help decision making about treatment initiation for men at this disease stage. Australian National Health and Medical Research Council and Cancer Councils, The Royal Australian and New Zealand College of Radiologists, Mayne Pharma Australia, Tolmar Australia.