The effects of various light power densities (milliwatts per centimeter of diffusing fiber [mW/cf]) and light doses (joules per centimeter of diffusing fiber [J/cf]) on the effectiveness of photodynamic therapy to endobronchial and tracheal tumors were evaluated at 46 different sites. All patients had squamous cell carcinoma or adenocarcinoma. They received 2 mg/kg body weight dihematoporphyrin ether intravenously 2 days before treatment bronchoscopy. Only one light treatment was delivered to the site. Patients were treated with diffusing cylinder light tips and underwent toilet bronchoscopy 2 days after photodynamic therapy. The percentage of obstruction was estimated before and after treatment and before and after toilet bronchoscopy. There was no difference between the effects resulting from power densities of 400 and 500 mW/cf, nor were there differences in the reactions between squamous cell carcinoma and adenocarcinoma. The amount of tumor that could be removed at the end of the treatment bronchoscopy, the amount of reobstruction by secretions and exudate seen at toilet bronchoscopy, and the final percent decrease in obstruction at the end of toilet bronchoscopy were proportional to the light dose. Because the final percentage decrease in obstruction plateaued at light doses of 400 to 500 J/cf and there was no statistically significant difference between 400 and 500 J/cf, we recommend using a power density of 500 mW/cf and a light dose of 400 J/cf during photodynamic therapy.