Background/objectives: Debilitating abdominal pain is a common symptom affecting patients with chronic pancreatitis (CP). CP pain is dynamic due to multiple underlying mechanisms. The objective of this study was to 1) evaluate changes in pain phenotype at one year follow-up and 2) validate putative pain biomarkers in a prospective cohort study. Methods: The Neuropathic and Nociceptive PROMIS-PQ questionnaires were used to classify pain for participants with in the PROCEED study. Putative serum biomarkers were measured via immunoassay. Results: At enrollment, 17.6 % (120/681) subjects with CP reported no pain in the previous year. Of those, 29 % experienced pain during the 1 yr follow-up whereas 18 % of those with pain prior to enrollment reported no pain during the 1 yr follow-up period. Of the 393 subjects with PROMIS-PQ data at enrollment, 212 also had follow-up data at 1 yr. Approximately half (53.3 %) of those individuals changed pain phenotype between baseline and follow-up. At 1 yr, serum TGFβ1 level was negatively correlated with nociceptive T-scores (p = 0.006). GP130 was significantly correlated with both nociceptive (p = 0.012) and neuropathic T-scores (p = 0.043) at 1 yr, which is consistent with the previously published findings. Conclusions: The positive association between TGFβ1 and pain is not maintained over time, suggesting it is a poor pain biomarker. However, serum GP130 is a consistent biomarker for mixed-type pain in CP. Preclinical studies show that targeting TGFβ1 or IL-6 (ligand for GP130) is sufficient to inhibit CP pain supporting further investigation of this as a potential therapeutic target.