It has been hypothesized that pro-inflammatory cytokines may play a pathogenic role in Alzheimer's disease (AD), and that n-3 polyunsaturated fatty acids may be protective against the development and progression of this disease. A reduced release of inflammatory cytokines by lipopolysaccharide (LPS)-stimulated peripheral blood mononuclear cells (PBMCs) from AD patients dietary supplemented with a mixture of eicosapentaenoic (EPA) and docosahexaenoic acid (DHA) was recently reported. On this basis, we investigated the possible differential effects of the two purified fatty acids on inflammatory cytokine release, a subject still not explored, even though of great pharmacological interest. We treated in vitro phytohaemagglutinin (PHA)- or LPS-stimulated PBMCs from AD patients and age-matched healthy controls (HCs) with purified EPA or DHA. Higher pro- to anti-inflammatory cytokine ratios, indicative of a pro-inflammatory profile, were observed in PHA-stimulated PBMCs from AD patients in basal conditions. The addition of both EPA and DHA markedly reduced the cytokine release, with DHA showing always a more prominent effect than EPA. However, whereas DHA reduced only the high IL-1β/IL-10 ratio, EPA was able to reduce also the IL-6/IL-10 ratio. In stimulated PMBCs from HCs the reducing effect on cytokine release was not always observed, or observed at a lower degree. In conclusion, whereas DHA appeared more powerful in inhibiting each single inflammatory cytokine, the proinflammatory profile of the AD patients' cells was better reverted by EPA to a profile more similar to that found in HCs. A combination of both the fatty acids, seems to be still the best solution.