Pharmacokinetics of alpha-difluoromethylornithine (DFMO) in plasma and polyamine levels in urine and erythrocytes (RBC) of subjects considered to be at a higher-than-normal risk for developing cancer and receiving DFMO in a phase I chemoprevention trial were monitored over a period of 6 months at DFMO doses ranging from 200 to 6400 mg/m2/day. DFMO pharmacokinetics was linear and attained an average peak plasma concentration of 58 micrograms/ml and an average area under the concentration x time curve from 0 to 6 h of 240 micrograms/ml.h at an administered dose of 1600 mg/m2. Transient decreases in RBC polyamine levels were observed in only 3 of 22 subjects; all of the others showed an increase in the levels at some time during DFMO administration. In contrast to these findings, 17 of 22 subjects showed a decline in urinary polyamines; 10 of 22 showed this decline by the end of the first month and the remaining subjects during subsequent administration of the drug. One subject with familial polyposis who had high RBC and urinary polyamine levels prior to DFMO treatment showed a significant decline in urinary polyamines and responded to DFMO treatment with nearly complete resolution of the polyps in the rectal stump. Our results suggest that (a) DFMO concentrations achieved in this study are adequate to modulate polyamine pools as reflected by their reduced urinary excretion; (b) the red blood cell polyamines are not reliable indicators of DFMO activity; and (c) the modulation of polyamines occurs at doses of DFMO that are tolerated by a majority of the subjects.