3153 Background: To address the limitations of exploratory studies, we conducted IMPACT2. The primary endpoint was to compare the progression-free survival (PFS) between patients treated with matched targeted therapies (MTT) selected on the basis of tumor molecular alterations and those whose treatment was not selected on the basis of alteration analysis (non-targeted therapy, NTT). Methods: Patients with metastatic cancer and targetable alterations were randomized between the 2 arms when eligibility criteria were met (Part A, 5/2014-4/2017; sponsor, Foundation Medicine). In Part B (3/2019-9/2023; sponsor, Tempus), patients who declined randomization selected the treatment arm ( NCT02152254 ). PFS/OS analysis (Kaplan Meier, Cox model; reference group, comparison, NTT). All cases were presented at weekly Molecular Tumor Board meetings; initial plan, to randomize 300 of 1,362 enrolled patients. Results: Overall, 829 patients (Parts A, 391; B, 438); were enrolled (med. age 59.5 yrs, range 18-84; men 50%; med. No. of prior therapies 3, range 0-17; most common cancers: gastrointestinal 34%, head/neck 13.5%, gynecologic 10%, sarcoma 9.53%; most common genomic alterations RAS 38%, PI3K 31%, FGFR 28%). In Part A, 69 were randomized (MTT, n=35; NTT, n=34); Part B (MTT, n=11, NTT=13; Selected arm: n=16: MTT, n=13; NTT, n=3). Results are shown (Table). MTT group: median PFS by therapy type: FDA approved/off-label, 4.7 months (95% CI, 2.93, 7.1); investigational, 2.6 months (95% CI, 2.27,4.87), combined FDA approved/investigational, 1.7 months (1.7, NR) (p=0.22). Conclusions: Considering the challenges to randomize patients (doi: 10.1038/s41698-022-00317-0), and that the study was underpowered for its original goal, no differences in response, PFS (primary endpoint), or OS was noted between the 2 arms, possibly because of the advanced stage disease and the complexity of tumor biology that cannot be addressed by randomization. Clinical trial information: NCT02152254 . [Table: see text]