An N-acetylcysteine (NAC)/oltipraz (OLZ) combination was studied in healthy volunteer smokers who received daily NAC (1200 mg/day) and were randomized to weekly placebo (Arm A), OLZ 200 mg (Arm B), or 400 mg (Arm C). Treatment was for 12 weeks with follow-up at 16 weeks. The objective was to study toxicity and the modulation of pharmacodynamic end points. After treatment of 19 of a planned 60 subjects, (Arm A, six; Arm B, four; and Arm C, nine), the study was closed because of toxicity. Eight subjects failed to complete 12 weeks of drug administration, (Arm A, two, and Arm C, six). The most frequent side effects were gastrointestinal, fatigue, conjunctival irritation, and skin rash. Pharmacodynamic end points were measured pretreatment and 48 h after the dose of OLZ at weeks 1, 5, and 12 and 4 weeks after the end of treatment. Glutathione (GSH) was measured in plasma and in peripheral blood lymphocytes (PBLs). Other end points measured in PBLs were the enzyme activities of total glutathione-S-transferase (GST), GSTpi, and NAD(P)H:quinone oxidoreductase; and the mRNA expression of gamma-glutamylcysteine synthetase gammaGCS), GSTpi, and NAD(P)H:quinone oxidoreductase. GSH in PBLs, GST (total), and the mRNA of gammaGCS showed increases at some time points in some subjects. Most consistent was the mRNA of gammaGCS, which showed a > or = 30% increase at one or more time points in 11 of 19 subjects. Other end points were unchanged. We concluded that NAC/OLZ modulates some end points related to GSH but is too toxic for chemoprevention at the doses used.