Clinical trial cost modeling is rarely done, but early protocol development and cost identification may reduce the uncertainty surrounding return on investment for therapies such as a new pulmonary TB drug. OBJECTIVE: To develop a Phase III trial protocol for a potential new pulmonary TB drug and estimate the trial costs. METHODS: We developed a trial protocol based on the study design submitted to the Food and Drug Administration (FDA) for rifapentine and the Good Clinical Practice (GCP) standards of the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Clinical trial experts reviewed the protocol and costs were assessed based on the Resource Based Relative Value Scale (RBRVS), expert opinion, and published prices for materials and services. We constructed a model that summarized costs per patient, per trial site, and per trial to implement the protocol. Entering the number of patients to be assessed, number of trial sites, number of trials, and anticipated drop out rates generated estimated costs. The model calculated the cost for all patients to complete the trial, costs associated with screening patients who drop out immediately, and savings that may be realized from patients dropping out during the trial. These figures generated lower and upper trial cost estimates. A 15% drop out rate at screening and a 15% drop out rate during the trial was assumed. RESULTS: For 850 to 1,000 patients to complete a Phase III trial 1,176 patients will need to be screened. The estimated total cost (2000 US dollars) is between $22.3 million and $22.6 million. CONCLUSIONS: A Phase III clinical trial for a new pulmonary TB drug will cost between $22.3 million and $22.6 million, assuming a drop out rate no greater than 15% at screening and 15% over the course of the trial.