Abstract Background: Aromatase inhibitors (AIs) are the standard adjuvant treatment of hormone-sensitive breast cancer (BC) in postmenopausal women. However, these therapies are associated with musculoskeletal complaints which may lead to non-adherence and early discontinuation. The aim of this study was to characterize the natural history of the AI-induced arthralgia. Methods: Postmenopausal women with stage I-III BC were prospectively enrolled at the onset of adjuvant AI therapy. Subjects completed the Brief Pain Inventory (BPI) questionnaire at baseline, 3, 6, 9, and 12 months. BPI worst pain scores were categorized as 0-3, 4-6 and 7-8. Multiple logistic regression analysis was used to evaluate the association between baseline factors and having a 2-point worsening in BPI worst pain score. A Cox-proportional hazards model was used to evaluate factors that influenced the time to first 2-point worsening in pain score. Clinically significant change was defined as ≥2-point increase from baseline. Those with a baseline BPI worst pain score of ≥9 were removed from analysis. Results: Among 180 consented subjects, 1 was found to be ineligible due to being perimenopausal, 42 subjects were lost to follow up, 17 subjects came off their AI and 8 subjects dropped out. Mean age was 61; 60% were white, 32% Black, 10%, Asian and 31% were Hispanic; 86% started on anastrazole; 24% had a change in AI. At baseline, 76 women had a BPI worst pain score between 0-3; 28 between 4-6; and 18 had 7+. Seventy subjects (64%) experienced at least a 2-point worsening of BPI from baseline and women who developed a 2-point worsening had a lower mean baseline BPI (2.57 vs. 3.75) compared to those who did not. Those experiencing an initial worsening in the first 6 months of therapy improved over time; while the patients experiencing BPI worsening at 9 months, continued to have progressive increase in BPI. In a multiple logistic regression model adjusting for AI switching, BMI, age, baseline score, prior chemotherapy, osteoarthritis, and prior hormone replacement therapy, those who had a baseline worst pain score between 4-6 (p=0.04) or 7+ (p=0.005) were less likely to develop a 2-point worsening in BPI at 12 months from baseline. Similarly, those with a baseline worst pain categorization of 7+ had a hazard ratio of 0.34 for time to 2-point worsening. Conclusions: Women with low baseline BPI score are more likely to develop worsening BPI score over time. Women who develop symptoms later in the course of their therapy are at greatest risk for persistent symptoms. This has implications for studies evaluating interventions for prevention of AI arthralgias. BPI mean scorePatients with >2 point increase in BPI at any timePatients with >2 point increase in BPI at 3 monthsPatients with >2 point increase in BPI at 6 monthsPatients with >2 point increase in BPI at 9 monthsPatients with >2 point increase in BPI at 12 monthsBaseline score3.752.57*2.592.302.11*1.85*Three month score4.254.526.08*4.684.133.87Six month score4.504.83*5.32*6.37*5.074.22Nine month score4.005.105.115.135.75*5.04Twelve month score4.255.37*5.484.865.666.21**p value <0.05 compared to baseline Citation Format: Lea Baer, Katherine D Crew, Danielle Awad, Kevin Kalinsky, Matt Maurer, Dawn L Hershman. Baseline joint pain predicts severity of subsequent joint symptoms in women initiating aromatase inhibitors for early stage breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P5-15-03.