Both oxidative damage and inflammation are elevated in brains of Alzheimer's disease (AD) patients, but their pathogenic significance remains unclear. The reduced AD risk associated with high intake of both nonsteroidal anti-inflammatory drugs (NSAIDs) and antioxidants suggests causal roles, but clinical trials in AD patients have yielded only limited or negative results. To test the potential efficacy and mechanisms of candidate approaches, we have explored conventional and unconventional NSAIDs, antioxidants, and combined NSAID/antioxidants in cell culture and animal models for AD (including aging APPsw transgenic mice and soluble Abeta rodent infusion models). The conventional NSAID ibuprofen has the strongest epidemiological support. At sustainable doses designed to mimic protective consumption in the epidemiology, ibuprofen reduces amyloid accumulation but suppresses a surprisingly limited subset of inflammatory markers in APPsw transgenic mice. Both Ab production (APP, beta- and gamma-secretases) and post-production pathways (those affecting Abeta aggregation or clearance: e.g., IL-1 or alpha1ACT) are potentially involved in ibuprofen and other NSAID anti-AD activities. The post-production pathways are predictably shared with other seemingly protective NSAIDs, including naproxen that do not lower Abeta42 in vitro. Using clinically feasible dosing, brain levels of NSAIDs appear too low to implicate a number of pharmacological dose targets that have been demonstrated in vitro. Ibuprofen did not suppress microglial markers related to phagocytosis. The putative anti-inflammatory omega-3 fatty acid DHA had a profound impact on pathogenesis but did not lower inflammation, while vitamin E was surprisingly ineffective in reducing oxidative damage or amyloid in the aged APPsw mouse. In contrast, the unconventional NSAID/antioxidant curcumin was effective, lowering oxidative damage, cognitive deficits, synaptic marker loss, and amyloid deposition. Curcumin proved to be immunomodulatory, simultaneously inhibiting cytokine and microglial activation indices related to neurotoxicity, but increasing an index of phagocytosis. Curcumin directly targeted Abeta and was also effective in other models, warranting further preclinical and clinical exploration.