Background: Sperm selection for intracytoplasmic sperm injection (ICSI) has traditionally relied on standardised methods of sperm processing combined with subjective sperm selection (motility/morphology). In 2012, live birth rates (LBRs) stood at ≈24% per cycle started (32% per cycle reaching embryo transfer). Objective(s):The main clinical objective was to determine the benefits of a hyaluronan (HA)-based sperm selection process for physiological intracytoplasmic sperm injection (PICSI). A parallel, mechanistic objective evaluated sperm chromatin integrity and the potential of PICSI to compensate for poor sperm quality. Design: A Phase III, parallel-arm, blinded randomised controlled trial (RCT) of efficacy of PICSI versus ICSI alongside mechanistic evaluation. Setting: The RCT ran from February 2014 to August 2016, involving NHS (n = 14) and private (n = 2) UK hospital or satellite clinics. Mechanistic work was conducted in three university-based research laboratories and a partnering small–medium enterprise from June 2015 to December 2017. Participants: Couples undergoing an ICSI procedure using freshly ejaculated sperm with female partners aged between 18 and 43 years and male partners aged between 18 and 55 years. Intervention: Health and Care Professions Council-registered embryologists used the Medicines and Healthcare products Regulatory Agency-registered (HA-coated) PICSI™ dish (Origio, Måløv, Denmark) to select a single sperm for injection. Control couples received standard care. Main outcome measures: Clinical – the primary outcome was full-term live birth (≥ 37 weeks’ gestation). Secondary outcome measures were confirmed clinical pregnancy (CP), miscarriage following confirmation and preterm live birth (< 37 weeks’ gestation). Mechanistic – measurement models were designed for deoxyribonucleic acid (DNA) fragmentation, compaction and HA binding [HA binding score (HBS)]. Results: A total of 2772 couples were randomised and 2752 couples were included in the primary analysis (PICSI, n = 1371; and ICSI, n = 1381). Clinical – primary outcome: 379 out of 1381 (27.4% PICSI) and 346 out of 1371 (25.2% ICSI) couples who were randomised (up to 24 hours before treatment) into the trial achieved a term live birth ≥ 37 weeks’ gestation [odds ratio (OR) 1.12, 95% confidence interval (CI) 0.94 to 1.34; p = 0.18]. Subgroup analyses did not reveal differences in treatment effects for HBS, maternal age, previous miscarriage, follicle-stimulating hormone or anti-Müllerian hormone levels and paternal sperm concentrations. Secondary outcomes: CP was achieved for 487 out of 1382 (35.2% PICSI) and 491 out of 1375 (35.7%, ICSI) couples (OR 0.98, 95% CI 0.84 to 1.15; p = 0.80). Miscarriage affected 60 out of 1381 (4.3% PICSI) and 96 out of 1371 (7.0% ICSI) of couples (OR 0.61, 95% CI 0.43 to 0.84; p = 0.003). Preterm LBRs were 46 out of 1381 (3.3% PICSI) and 45 out of 1371 (3.3% ICSI) (OR 1.02, 95% CI 0.67 to 1.55; p = 0.94). Mechanistic: in the subset of samples examined, HBS correlated with sperm motility, concentration, fertilisation rate and DNA fragmentation. Sperm DNA compaction was weakly associated with clinical pregnancy rates (CPRs), but neither HBS nor DNA fragmentation was predictive of any clinical outcome. Limitations: Embryologists were not blinded and limited data were available from poorer samples and non-random sample selection in the mechanistic cohort. Prepared rather than raw semen was used for tests of DNA integrity. Conclusions: PICSI offered no clear advantage in relation to the primary outcome. PICSI led to a reduced miscarriage risk, but had no effect on CPR or preterm LBR. Future work: Re-evaluate PICSI focusing on CP and miscarriage rates and consider aspects of sperm quality that PICSI favours. Trial registration: Current Controlled Trials ISRCTN99214271.