Twenty-eight porphyrins were evaluated for tumor localization as delineated by fluorescence using a transplantable KHJJ mammary carcinoma in BALB/c mice as the tumor model. Five of the 28 porphyrins were found to localize and of these, one, i.e., uroporphyrin I (UROP I), showed a higher tumor:skin ratio than any of the others; moreover, as no measurable UROP I was present in the gut, the tumor:intestinal porphyrin ratio under the conditions of assay was infinity. Because hematoporphyrin derivative (HPD), a complex mixture of porphyrins has been studied extensively as a tumor localizer, we compared HPD with UROP I at differing doses (2-40 mg/kg) and at different times (3-96 h) following i.v. administration. Dose response curves showed tissue levels of porphyrin to plateau out at doses above 20 mg/kg. Peak tumor HPD and UROP I levels attained 6-18 h after i.v. administration (40 mg porphyrin/kg) were comparable, but tumor retention of HPD over the ensuing 96 h was higher. The ratio of UROP I in tumor compared to skin was significantly greater throughout the period of observation. At all times, no UROP I was detectable in gastrointestinal mucosa. At differing doses (10-40 mg/kg), the tumor:skin ratio for HPD ranged from 1.47-1.85, and for UROP I from 6.06-12.33. As a function of time (6-72 h), the tumor:skin ratios respectively were 1.03-2.38, and 11.9 to infinity. At all times, the tumor:colon mucosa ratio at different doses for HPD approached 1 and for UROP I was infinity. We conclude that the greater specificity of tumor uptake by UROP I and its lack of retention by gut mucosa warrants further study to determine its potential clinical application as a diagnostic marker, particularly for early mucosal cancer, and in photoradiation therapy.