To evaluate the association of vitamin D level with incident rheumatoid arthritis (RA) in a patient population using electronic health records (EHR). Case-control study with data extracted from EHR from 1/1/2001 to 12/31/2012 in the Geisinger Health System (GHS). Incident RA was defined as International Classification of Disease-9 code 714.0 twice by a GHS rheumatologist. Patients were identified at time of RA diagnoses and were matched 1:5 for age and gender with non-RA controls. Vitamin D levels were identified and extracted prior to RA diagnosis. The subjects were followed retrospectively with regard to their vitamin D levels; the most recent value of vitamin D prior to the RA diagnosis was used in the analysis. Vitamin D levels were treated both as continuous and categorical with two different cutoff values, 30 and 20 ng/ml. The association between vitamin D and RA was presented as the odds ratios with 95 % confidence intervals (OR, 95 % CI) from a conditional logistic regression model adjusting for obesity and smoking status. A total of 270 patients with incident RA and 1,341 matched controls were identified. The RA patients were 83.3 % female with median age at RA diagnosis of 62 years. The adjusted OR (95 % CI) for the association of vitamin D levels with incident RA compared with controls was 1.00 (0.99, 1.01), 0.98 (0.75, 1.29) and 1.12 (0.80, 1.57) for continuous, <30 and <20 ng/ml vitamin D levels, respectively. Subgroup analysis according to gender or rheumatoid factor positivity yielded similar results. In this patient population, vitamin D levels were not associated with the development of RA.