A defined mixture of rho iso-alpha-acids (RIAA), a modified hop extract, was evaluated for anti-inflammatory efficacy and safety. RIAA inhibited LPS-stimulated PGE(2) formation with >200-fold selectivity of COX-2 (IC(50)=1.3 microg/ml) over COX-1 (IC(50)>289 microg/ml). This occurred only when RIAA was added prior to, but not post, LPS stimulation. Consistent with this observation, RIAA produced no physiologically relevant, direct inhibition of COX-1 or COX-2 peroxidase activity. This suggests that RIAA inhibits inducible but not constitutive COX-2. In support, we found RIAA showed minimal PGE(2) inhibition (IC(50)=21mug/ml) relative to celecoxib (IC(50)=0.024 microg/ml), aspirin (IC(50)=0.52 microg/ml) or ibuprofen (IC(50)=0.57 microg/ml) in the AGS gastric mucosal model, where COX-1 and -2 are expressed constitutively. Taken together these results predict RIAA may have lower potential for gastrointestinal and cardiovascular toxicity observed with COX enzyme inhibitors. Following confirmation of bioavailable RIAA administered orally, gastrointestinal safety was assessed using the fecal calprotectin biomarker in a 14-day human clinical study; RIAA (900 mg/day) produced no change compared to naproxen (1000 mg/day), which increased fecal calprotectin 200%. Cardiovascular safety was addressed by PGI-M measurements where RIAA (1000 mg) did not reduce PGI-M or affect the urinary PGI-M/TXB(2) ratio. Drug interaction potential was evaluated against six major CYPs; of relevance, RIAA inhibited CYP2C9. Toxicity was assessed in a 21-day oral, mouse subchronic toxicity study where no dose dependent histopathological effects were noted. Clinically, RIAA (1000 mg/day) produced a 54% reduction in WOMAC Global scores in a 6-week, open-label trial of human subjects exhibiting knee osteoarthritis.