Dyspepsia patterns during the course of duodenal ulcer (DU) were examined in terms of the number of consecutive months that patients experienced dyspepsia either intermittently or continuously, or were dyspepsia-free. The patients (n = 224) were community-based. After ulcer healing, they were evaluated every 3 months for less than or equal to 7 years. Antiulcer agents were prescribed at the physicians' discretion. We analyzed the data on a patient-month basis using a 2-state Markov chain model. The fitted mean number of consecutive months with dyspepsia and the number of months dyspepsia-free were, respectively, 1.7 and 14 in patients on antiulcer agents, and 1.5 and 9 in patients not taking them. On average, patients experienced dyspepsia during 15% of months studied if they were on antiulcer agents, and during 20% of months if not. Regardless of antiulcer therapy, dyspepsia periods were lengthened by age, especially greater than 60 years; male sex; widowhood/divorce/separation; increasing acetaminophen use; and increasing length of time since initial onset of ulcerlike dyspepsia. Dyspepsia-free periods were lengthened by age greater than 70 years, and shortened by widowhood/divorce/separation and increasing cigarette consumption. These results suggest that dyspepsia occurs for less than or equal to 20% of the time in DU patients on about a yearly basis. The least dyspepsia may be expected in patients neither widowed, divorced, nor separated, in whom length of time since the initial onset of ulcerlike dyspepsia does not exceed 7 years, and who neither smoke nor use analgesics regularly.