Prolonged use of non-steroidal anti-inflammatory drugs (NSAIDs) may reduce the risk of developing Alzheimer's disease (AD) and delay disease onset. Negative results of clinical AD trials were rationalised by the discovery that certain NSAIDs reduce amyloid-beta( 1-42) (A beta(1- 42)) peptide production, the proposed central culprit in AD pathophysiology and main constituent of amyloid plaques, whereas other compounds do not affect A beta levels. Latter observations motivated further in-vitro and in-vivo research regarding the applicability of NSAIDs in treating and/or preventing AD. We used the age-dependent cognitive decline in the APP23 transgenic mouse model for AD to evaluate disease-modifying efficacy of chronic ibuprofen treatment at the cognitive level. At age 6 weeks, heterozygous APP23 mice and control littermates were subcutaneously implanted with osmotic pumps delivering saline or ibuprofen (50 mg/kg daily). After 2 months of treatment, a 3-week washout period prevented bias from potential symptomatic effects before cognitive evaluation commenced. Ibuprofen-treated APP23 mice performed significantly better than their sham-treated counterparts and almost attained the same level of performance as control animals on a complex visual-spatial learning task. This study clearly reports disease-modifying efficacy of ibuprofen at the cognitive level in transgenic mice modelling AD.