12020 Background: Early referral to palliative care was associated with improved health-related quality of life (HRQL) and overall survival (OS) in a US phase 3 trial in lung cancer patients (pts). International studies in mixed cancer types have reported conflicting results. PEARL aimed to determine whether early referral to palliative care would improve HRQL, OS, and resource use in Australian pts with recently diagnosed, advanced thoracic cancers. Methods: Eligible participants (pts) in this unblinded, multi-centre, randomised, phase 3 trial had advanced thoracic cancers diagnosed within 60 days, and the ability to complete patient-rated outcome measures (PROMS). Pts were randomly allocated to early referral to palliative care (ER) or referral at clinician’s discretion (DR). All pts received standard oncological care. PROMS were completed at baseline, every 3-4 weeks for 6 months, then 6-8 weekly. The primary objective was to determine the frequency of sustained, substantial improvements in HRQL, defined as a 5-point improvement in the FACT-L Trial Outcome Index (TOI) maintained for at least 2 consecutive assessments. Secondary outcomes included OS, documentation of advanced care plan (ACP), PROM scores at 12 weeks, anxiety/depression (PROMIS-ED), lung cancer symptoms (FACT-L), global HRQL (ICECAP-SCM), carer-satisfaction and burden, and understanding of illness and prognosis. The accrual target of 200 gave 80% power (alpha 0.05) to detect an absolute improvement of 20% in the proportion of pts achieving the primary endpoint. Results: 113 pts and 78 carers were recruited when the trial closed for slow accrual. Pt characteristics were well balanced: 88 (75%) had NSCLC, 18 (16%) small cell and 7 (6%) mesothelioma. Median age was 69 (IQR 62-74), 63 (56%) were male; systemic anti-cancer therapy ongoing or planned in 88 (78%). Median follow-up was 30 months. First consultations with a palliative care specialist within 60 days of diagnosis occurred in more pts assigned ER vs DR (57% vs 3.5%). Sustained substantial improvements in FACT-L TOI were reported by similar numbers of pts assigned ER vs DR: 33% vs 32%, p = 0.9. OS was similar among those assigned ER versus DR (median 12 vs 18.4 months, p = 0.11). A similar % had a written advanced care plan at death: 15/40 (39%) vs 15/33 (47%). We found no important differences between arms in global HRQL (ICECAP-SCM), depression/anxiety (PROMIS-ED), lung cancer symptoms (FACT-L), carer satisfaction (FAMCARE-2), carer burden (CRA), or understanding of illness by carers or pts. Conclusions: Early referral to palliative care, compared with discretionary referral, did not improve important outcomes for Australian thoracic cancer pts or carers. Our findings suggest that the palliative care needs of such pts were addressed equally well by delayed referral when clinically indicated, resulting in reduced burden for resource-limited specialist palliative services. Clinical trial information: ACTRN12617000166370.