Aromatase inhibitor (AI)-associated arthralgia limits adherence to therapy in breast cancer. The pathophysiology may involve vitamin D status. We wished to establish the optimal concentration of 25(OH)D that prevents or minimizes arthralgia. We used a prospective cohort of 290 women starting AI in whom baseline vitamin D was measured. All received daily vitamin D(3) (800 IU) with calcium. Women with baseline 25(OH)D concentration <30 ng/ml also received 16,000 IU of D(3) orally every 2 weeks. The primary outcome was incident or worsening joint pain derived from baseline and 3-month visual analogic scale (VAS) for joint pain. Regression models were used to analyse the association between vitamin D concentrations at 3 months and pain adjusting for age, BMI, season when the sample was drawn, aromatase inhibitor (exemestane vs. letrozole/anastrozole), prior tamoxifen therapy, baseline NTX, and previous fracture. 90% of women had a 25(OH)D <30 ng/ml at baseline. After supplementation (daily 800 IU and additional 16,000 IU every 2 weeks), 50% of them still failed to reach adequate concentrations at 3 months. In the whole cohort, there was an increase in joint pain (mean 1.16 points SD 2.66; P < 0.001) and the increase was significantly (P = 0.02) attenuated in those that reached concentrations of 25(OH)D of ≥40 ng/ml, with a lower risk of incident arthralgia (OR 0.12 ** [0.03 to 0.40]). A target concentration of 40 ng/ml 25OHD may prevent development of AI arthralgia but higher loading doses are required to attain this level in women with deficiency at baseline.