OBJECTIVE: This study examined the association between access to health care and three classifications of diabetes status: diagnosed, undiagnosed, and no diabetes. RESEARCH DESIGN AND METHODS: Using data from the 1999-2004 National Health and Nutrition Examination Survey, we identified 110 "missed patients" (fasting plasma glucose >125 mg/dl but without diagnoses of diabetes), 704 patients with diagnosed diabetes, and 4,782 people without diabetes among adults aged 18-64 years. The population percentage undetected among adults with diabetes and the odds ratio of being undetected among adults who reported not having diabetes were compared between groups based on their access to health care. RESULTS: Among those with diabetes, the percentages having undetected diabetes were 42.2% (95% CI 36.7-47.7) among the uninsured, 25.9% (22.9-28.9) among the insured, 49.3% (43.0-55.6) for those uninsured >1 year, 38.7% (29.2-48.2) for those uninsured <or=1 year, and 24.5% (21.7-27.3) for those continuously insured over the past year. Type of insurance, number of times receiving health care in the past year, and routine patterns of health care utilization were also associated with undetected diabetes. Multivariate adjustment indicated that having undetected diabetes was associated with being uninsured (odds ratio 1.7 [95% CI 1.0-2.9]) and with being uninsured >1 year (2.6 [1.4-5.0]). CONCLUSIONS: Limited access to health care, especially being uninsured and going without insurance for a long period, was significantly associated with being a "missed patient" with diabetes. Efforts to increase detection of diabetes may need to address issues of access to care.