Purpose To determine the longevity of vitamin D status following cessation of vitamin D3 supplementation, 2 and 3 years after a 1 year randomised double blind placebo controlled trial: (Vitamin D and Cardiovascular Risk (VICtORY)); and to investigate possible predictive factors. Method Of the 305 Caucasian non-smoking postmenopausal women randomised to ViCtORY (2009-2010), participants who had not taken vitamin D supplements since the trial ended were invited to attend follow up visits. Total 25-hydroxyvitamin D (25OHD) and 24,25-dihydroxyvitamin D (24,25OH2D) were measured by dual tandem mass spectrometry of serum samples following removal of protein and de-lipidation; the original RCT samples were re-analysed simultaneously. Vitamin D binding protein (VDBP) was measured by monoclonal immunoassay. Results In March 2012 and March 2013, 159 women (mean (SD) age 67.6 (2.1) years) re-attended, distributed between the original treatment groups: daily vitamin D3 400IU; 1000IU; and placebo. One month after the RCT ended (March 2010) the proportion of women in placebo, 400IU, and 1000IU vitamin D3 groups, respectively, with 25OHD<25 nmol/ L was 15%,0%, 0% (Chi-square p<0.001, n=46,44,54). After 2 years (March 2012) it was 22%, 4%, 4% (p=0.002, n=50,48,57); after 3 years 23%, 13%, 15% (p=0.429,n=48,45,52). The respective proportion of women with 24,25OH2D < 2.2 nmol/L were 50%, 2%, 2% (1 month, p>0.001, n=46,44,54); 42%, 33%, 12% (2y, p=0.002,n=50,48,57) and 45%, 27%, 29% (3y, p=0.138, n=47,45,51,). VDBP was a predictor of circulating 25OHD longevity (beta for VDBP in µg/ml:0.736; 95% CI 0.216-1.255,p=0.006) but not 24,25OH2D.