Laboratory-based surveillance is a foundation for public health and is essential for determining the incidence of most foodborne diseases caused by bacterial pathogens; however, reported cases represent a subset of infections in the community. To identify the factors associated with seeking medical care and submitting a stool specimen among persons with acute diarrheal illness, we used multivariate logistic regression to analyze data from two 12- month population-based telephone surveys conducted in the Foodborne Diseases Active Surveillance Network (FoodNet) from 2000 to 2003. Of 31,082 persons interviewed, 5% reported an acute diarrheal illness in the four weeks prior to the interview; of these, 20% sought medical care. On multivariate analysis, among persons with an acute diarrheal illness, factors associated with seeking medical care included: male sex; age <5 or >or=65 years; household income <25,000 dollars; having health insurance; diarrhea duration >or=3 days; having bloody diarrhea, fever, vomiting, sore throat, or cough. Of those seeking medical care, 19% provided a stool sample. Bloody diarrhea (odds ratio [OR] 3.35; 95% confidence interval [CI] 1.18-9.51) and diarrhea duration >or=3 days (OR 3.81; 95% CI: 1.50-9.69) were the most important factors associated with submission of a stool specimen. Cases of acute diarrheal illness ascertained through laboratory-based public health surveillance are likely to differ systematically from unreported cases and likely over-represent those with bloody diarrhea and longer diarrhea duration.