A two-step Phase I study of piroxicam (PXM) and a-difluoromethylornithine (DFMO) alone and in combination was initiated to assess toxicity and the impact of these drugs on several biological markers. In step 1, 12 subjects with a history of skin cancers were assigned to receive PXM 10 mg every day (q.d.) or 10 mg every other day (q.o.d.). The dosage of PXM 10 mg q.o.d. was tolerated. No changes were seen in 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced ornithine decarboxylase (ODC) or urinary polyamine levels. Steady-state serum levels of PXM were consistent with the oral dose level. In step 2, 31 subjects with stage 0 or I nonmelanoma skin cancers, stage A or B prostate or colon cancer, or stage I breast cancer or who had a family history of cancer were randomized to receive DFMO 0.5 g/m2, PXM 10 mg q.o.d., or the combination of DFMO and PXM. In addition to the biological markers of TPA-induced ODC activity in skin biopsies and urinary polyamine levels, we measured urinary 11-dehydrothromboxane B2, a specific metabolite of thromboxane A2. Of the 12 subjects on DFMO/PXM, 2 dropped out for non-drug-related reasons. Three developed grade-2 drug-related toxicities. One subject developed dyspnea that resolved and was able to continue on the study for 6 months. One subject who developed diarrhea that resolved after 5 days was also able to restart the drug without a recurrence. A third subject described intermittent episodes of tinnitus starting 4 h after taking PXM that lasted only 5 s and did not progress on treatment. Comparing the 6-month measurements with pretreatment, DFMO/PXM or DFMO significantly reduced TPA-induced ODC levels (Ps, 0.03 and 0.05). Urinary polyamine levels of spermidine decreased slightly with the DFMO/PXM or DFMO alone, whereas putrescine decreased with PXM alone. Levels of 11-dehydrothromboxane B2 were depressed by PXM and PXM/DFMO. The doses of DFMO/PXM determined in step 2 are potential starting dosages for Phase IIa and IIb chemoprevention trials.