The chemopreventive efficacy of Targretin was evaluated in various rodent cancer models. In the rat model of 4-hydroxybutyl(butyl)nitrosamine (OH-BBN)-induced urinary bladder cancer, it was found that Targretin administered in the diet (beginning one week after the last OH-BBN treatment) for 5.5 months increased the number and size of urinary bladder cancers. In the azoxymethane (AOM)-induced model of colon carcinogenesis (in which rats develop minimally invasive colonic cancers), Targretin was ineffective as a chemopreventive agent, decreasing neither tumor incidence nor multiplicity. Treatment of Min mice with Targretin for 45 days similarly failed to decrease the multiplicity of small intestinal tumors. Similarly, no preventive efficacy was noted for Targretin when the incidence of tumors in the head and neck model (squamous cell tongue tumors) induced by 4-nitroquinoline 1-oxide (4-NQO) were examined. In contrast, use of even a suboptimal dose of Targretin (40 ppm) in a sensitive breast cancer model [methylnitrosourea (MNU)-induced ER+ mammary cancers] reduced cancer multiplicity by 60%. Finally, based on the hypothesis that Targretin may decrease the expression of COX‑2, the effects of Targretin and COX inhibitors were compared in these models. There was minimal overlap of efficacy. That is, models which were relatively susceptible to NSAIDs or COX-2 inhibitors tended not to be sensitive to Targretin and vice versa.