W e report data from a pilot study examining vitamin D deficiency in type 1 and type 2 diabetes. Serum 25-OH vitamin D (25-OH-D) levels were measured in type 1 (n 50) and type 2 diabetic (n 63) individuals at the time of their routine visits for diabetes management. Patients had no prior history of metabolic bone disease, vitamin D deficiency, parathyroid disease, malabsorption, or significant elevations of serum creatinine or liver enzymes. Most subjects (74.3%) were prescribed a daily multivitamin as part of their routine diabetes care, but none were taking additional supplements. Serum was obtained during “light” (April to September) and “dark” (October to March) months. Serum 25-OH-D levels were measured using a competitive protein-binding assay (Esoterix, Calabasas Hills, CA). Vitamin D levels were classified as deficient (0–20 ng/ml) or expected ( 20 ng/ml). Although 25-OH-D levels 32 ng/ml are optimal (1), levels 20 ng/ml are usually considered acceptable and are generally not treated in patients without metabolic bone disease. Type 1 diabetic individuals were significantly younger than those with type 2 diabetes ([mean SE] 49.0 1.7 vs. 61.2 1.5 years, P 0.001) and had a lower BMI (26.2 0.7 vs. 33.9 1.0 kg/m, P 0.001). The majority (63.5%) of the type 2 diabetic individuals were deficient compared with 36% of the type 1 diabetic patients. The unadjusted mean 25-OH-D level for type 2 diabetic patients was 17.1 1.2 ng/ml (P 0.05) compared with 23.6 1.2 ng/ml for type 1 diabetes (P 0.05). Of the patients who were deficient, most were prescribed multivitamins (72.2% of type 2 diabetes, 72.5% of type 1 diabetes). Most patients having serum 25-OH-D levels 20 ng/ml were also prescribed multivitamins (73.9% of type 2 diabetes, 78.1% type 1 diabetes). Serum was obtained with equivalent distribution in light and dark months among both groups. The 25-OH-D levels inversely correlated with BMI (P 0.01) but were not directly related to age or sex. ANCOVA was used to compare 25-OH-D levels in type 1 and type 2 diabetic patients while adjusting for BMI and age. As was noted in the unadjusted levels, significantly lower 25-OH-D levels were found in type 2 relative to type 1 diabetic patients (mean adjusted 25-OH-D 18.1 1.4 vs. 22.9 1.6 ng/ml, P 0.001). Among type 2 diabetic patients, mean 25-OH-D level was similar in those taking insulin compared with those on oral hypoglycemic agents (mean 25-OH-D 17.6 1.6 and 16.8 1.5 ng/ml, respectively; P 0.73). Our findings suggest that vitamin D deficiency is more common in type 2 diabetes than in type 1 diabetes, unrelated to age, sex, or insulin treatment. Although higher BMI was associated with lower 25OH-D levels, the difference in vitamin D levels between type 2 and type 1 diabetic individuals persisted after adjusting for BMI. This difference may be therapeutically significant in that mean vitamin D levels in type 2 diabetic patients were in the deficient range whereas mean levels in the type 1 diabetic patients reached the expected level. Despite most type 2 diabetic patients being prescribed a daily multivitamin, usually containing 400 IU of vitamin D, the percent who were deficient remained unchanged. Further studies are needed to better understand the causes and clinical significance of the observed hypovitaminosis D and to investigate response to vitamin D replacement therapy.