With 64,050 new diagnoses and 50,550 deaths in the US in 2023, pancreatic ductal adenocarcinoma (PDAC) is among the most lethal of all human malignancies. Early detection and improved prognostication remain critical unmet needs. We applied next-generation metabolomics, using quantitative tandem mass spectrometry on plasma, to develop biochemical signatures that identify PDAC. We first compared plasma from 10 PDAC patients to 169 samples from healthy controls. Using metabolomic algorithms and machine learning, we identified ratios that incorporate amino acids, biogenic amines, lysophosphatidylcholines, phosphatidylcholines and acylcarnitines that distinguished PDAC from normal controls. A confirmatory analysis then applied the algorithms to 30 PDACs compared with 60 age- and sex-matched controls. Metabolic signatures were then analyzed to compare survival, measured in months, from date of diagnosis to date of death that identified metabolite ratios that stratified PDACs into distinct survival groups. The results suggest that metabolic signatures could provide PDAC diagnoses earlier than tumor markers or radiographic measures and offer insights into disease severity that could allow more judicious use of therapy by stratifying patients into metabolic-risk subgroups.