Background: This post hoc nested case-control analysis of the TNT study was designed to investigate whether baseline vitamin D level is a significant predictor of cardiovascular risk among statin-treated patients and whether changes in vitamin D after treatment with atorvastatin are associated with improved cardiovascular outcomes. Methods: A total of 10,001 patients with stable coronary heart disease were randomized to atorvastatin 80 or 10 mg for a median of 4.9 years. This analysis included 1,509 patients (497 with a subsequent cardiovascular event and 1,012 without an event) with vitamin D levels determined at baseline and 1 year. Event rates were analyzed by Cox proportional hazard model by baseline vitamin D levels, with vitamin D as a continuous variable, and with change in vitamin D level as the predictor. Results: Vitamin D deficiency (<15 ng/mL) or insufficiency (15- <30 ng/mL) was present in 108 (7.2%) of 1,509 and 625 (41.4%) of 1,509 of patients, whereas 46 (3.0%) of 1,509 had elevated vitamin D. There was no relationship between baseline vitamin D levels or change in vitamin D levels and cardiovascular events or mortality. Modeling of events with vitamin D as a continuous variable similarly showed no relationship of vitamin D to events. These findings held true after adjustment for seasonal variations in vitamin D and other confounders. Conclusion: In statin-treated patients with stable coronary heart disease, vitamin D levels did not predict cardiovascular risk. Changes in plasma concentrations of vitamin D after 1 year of treatment made no contribution to the efficacy of atorvastatin therapy. © 2012 Mosby, Inc. All rights reserved.