Accumulation of amyloid beta-peptide (Abeta) is considered a key step in the etiology of Alzheimer's disease. Abeta is produced by sequential cleavage of the amyloid precursor protein by beta- and gamma-secretase enzymes. Consequently, inhibition of gamma-secretase provides a promising therapeutic approach to treat Alzheimer's disease. Preclinically, several gamma-secretase inhibitors have been shown to reduce plasma and brain Abeta, although they also produce mechanism-based side effects, including thymus atrophy and intestinal goblet cell hyperplasia. The present studies sought to establish an efficient screen for determining the therapeutic window of gamma-secretase inhibitors and to test various means of maximizing this window. Six-day oral administration of the gamma-secretase inhibitor N(2)-[(2S)-2-(3,5-difluorophenyl)-2-hydroxyethanoyl]-N(1)-[(7S)-5-methyl-6-oxo-6,7-dihydro-5H-dibenzo[b,d]azepin-7-yl]-l-alaninamide (LY411,575) reduced cortical Abeta(40) in young (preplaque) transgenic CRND8 mice (ED(50) approximately 0.6 mg/kg) and produced significant thymus atrophy and intestinal goblet cell hyperplasia at higher doses (>3 mg/kg). The therapeutic window was similar after oral and subcutaneous administration and in young and aged CRND8 mice. Both the thymus and intestinal side effects were reversible after a 2-week washout period. Three-week treatment with 1 mg/kg LY411,575 reduced cortical Abeta(40) by 69% without inducing intestinal effects, although a previously unreported change in coat color was observed. These studies demonstrate that the 3- to 5-fold therapeutic window for LY411,575 can be exploited to obtain reduction in Abeta levels without induction of intestinal side effects, that intermittent treatment could be used to mitigate side effects, and that a 6-day dosing paradigm can be used to rapidly determine the therapeutic window of novel gamma-secretase inhibitors.