Abstract Background: Per- and polyfluoroalkyl substances (PFAS) are environmentally persistent synthetic chemicals detectable in the blood of most Americans. Elevated exposures to perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA), the two most prevalent PFAS, have been associated with several health outcomes including metabolic outcomes and some cancers. To evaluate possible mechanisms related to the biological effects of these chemicals, we conducted a pooled metabolome-wide association study (MWAS) of circulating PFOS and PFOA among 3,647 participants included in one of eight nested case-control serum metabolomic profiling studies conducted in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) population. Methods: All metabolomic profiling was conducted by Metabolon Inc, using liquid-phase or gas chromatography coupled with mass spectrometry. All metabolites including PFOS and PFOA were centered (mean=0, standard deviation=1) and log-transformed. We conducted multivariable linear regression analyses to estimate study-specific associations between metabolite levels and PFOS and PFOA, adjusted for age at specimen collection, race/ethnicity, body mass index, smoking, case-control status, study center, and year of specimen collection. Sex was additionally adjusted for studies including both men and women. We then conducted combined the study-specific findings through meta-analysis using random effects models. Results: The meta-analysis results of 1,040 metabolites for PFOS and 1,103 for PFOA identified 83 and 67 metabolites, respectively, associated at Bonferroni-corrected significance thresholds (P=4.81 × 10−5 for PFOS; P=4.53 × 10−5 for PFOA). For both PFOS and PFOA, the strongest positive associations were observed for the xenobiotics 3,5-dichloro-2,6-dihydroxybenzoic acid [PFOS, summary beta coefficient =0.74, 95% confidence interval=0.67,0.80; PFOA, 0.48 (0.35,0.61)] and 3-bromo-5-chloro-2,6-dihydroxybenzoic acid [0.57 (0.43,0.72) and 0.39 (0.23,0.55)]. We also observed strong positive associations with PFOS and PFOA for metabolites related to the sphingomyelin (SM) pathway, such as hydroxypalmitoyl sphingomyelin (d18:1/16:0(OH)) [PFOS, 0.44 (0.24,0.64)] and SM (d18:1/18:0) [PFOS, 0.35 (0.24,0.46); PFOA, 0.25 (0.16,0.34)], and SM (d18:2/24:2) [PFOS, 0.32 (0.17,0.47)]. The PFOS metabolite associations remained after model adjustment for PFOA; however, the PFOA associations were greatly attenuated after controlling for PFOS. Conclusions: In this MWAS of PFAS, to our knowledge the largest to date, we observed robust positive associations with serum PFOS for several xenobiotic compounds and for metabolites related to the sphingomyelin pathway. Further investigation of these metabolites may offer insight into PFOS-related biologic effects contributing to disease development. Citation Format: Jongeun Rhee, Steven C. Moore, Demetrius Albanes, Tracy M. Layne, Erikka Loftfield, Rachael Stolzenberg-Solomon, Linda M. Liao, Mary C. Playdon, Mark P. Purdue. A metabolomic investigation of serum perfluorooctane sulfonate and perfluorooctanoate. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4223.