INTRODUCTION: It is known that the therapeutic use of drugs is often a major contributor to iatrogenic illness. Previous estimates of society's cost of drug-induced illness has been estimated to be between $30.1 billion to $136.8 billion. Therefore, the cause and prevention of iatrogenic illness is a topic of great importance and interest to clinicians, educators, administrators, and society. Unfortunately, previous studies have had key disadvantages. First, panels for achieving necessary probabilities have been limited in scope and sample size. Second, previous models have not addressed the recursive nature of health-related events. OBJECTIVES: This study had 2 objectives: (1) To test and evaluate a questionnaire that will be used in future studies to assess the probabilities of the various possible outcomes of drug related morbidity and mortality (DRMM); (2) To use this preliminary questionnaire data to produce a cost-of-illness model demonstrating the ability of Markov processes to model health economic questions. METHODS: The questionnaire was pre-tested on 22 physicians affiliated with UAMS health center. Of these, 11 (50%) responded to the survey. After revisions were made to this questionnaire it was mailed to 298 randomly selected physicians in the state of Arkansas. Of these, 58 usable questionnaires were obtained. Probabilities for the model were compiled based on this data. A Markov model was constructed using TreeAge Data software. Ten possible outcome states were identified for the model. Monte Carlo simulation was used to establish an expected value for the cost-of-illness related to DRMM. RESULTS: Extrapolations of the data indicate that the cost-of-illness in the United States related to DRMM is approximately $38 to $52 billion dollars.