Oral premalignancy serves as an ideal model for study of chemopreventive agents. Although 13-cis-retinoic acid showed reversal of oral premalignancy, toxicity, and reversal of clinical response after cessation of therapy obviated its widespread use. A search for nontoxic agents with cancer preventive activity led us to evaluate Bowman Birk Inhibitor (BBI) formulated as BBI Concentrate (BBIC). We previously reported encouraging results in a phase IIa trial of BBIC in patients with oral leukoplakia with measurable clinical responses and favorable biomarker changes. On the basis of these results, we undertook a randomized, placebo controlled phase IIb trial with patients receiving BBIC or placebo for 6 months, with assessment of clinical response and change in lesion area as primary end point and an intent-to-treat analysis. One hundred and thirty two subjects were randomized; and 89 subjects completed six months on study drug or placebo. Both placebo and BBIC showed a statistically significant decrease in mean lesion area of 17.1% and 20.6%, respectively, and partial or greater clinical responses of 30% and 28% respectively. No significant difference between placebo and study drug arms was observed. Histologic review, review of photographs of lesions, and comparison of serum neu protein and oral mucosal cell protease activity also did not show significant differences between study arms. Probable reasons for these negative results were considered, are discussed, and include a placebo with non-BBIC clinical activity and reduced pharmacokinetic availability of the second batch of BBIC. This experience should be a strong cautionary note to those considering "Green" chemoprevention.