Considering that several pathways leading to cell death are activated in cerebral ischemia, we tested in mouse models of transient and permanent ischemia a drug cocktail aiming at distinct pharmacological targets during the evolution of ischemic injury. It consists of minocycline--an antibiotic with anti-inflammatory properties, riluzole--a glutamate antagonist, and nimodipine--a blocker of voltage-gated calcium channels. Administered 2 h after transient or permanent MCAO, it significantly decreased the size of infarction, by approximately 65% after transient and approximately 35% after permanent ischemia and markedly improve clinical recovery of mice. In both experimental models a three-drug cocktail achieved significantly more efficient neuroprotection than any of the components tested alone. However, some interesting observation emerged from the single-drug studies. Treatment with minocycline alone was efficient in both experimental models while treatment with glutamate antagonist riluzole conferred neuroprotection only after transient MCAO. Immunohistochemical analysis following three-drug treatment revealed reduced microglia/macrophages and caspase-3 activation as well as preserved GFAP immunoreactivity following transient ischemia. No detectable differences in the levels of Mac-2, GFAP and caspase-3 immunoreactivities were observed 72 h after permanent MCAO. These marked differences in the brain tissue responses to ischemic injury and to treatments suggest that different pathological mechanisms may be operating in transient and permanent ischemia. However, the three-drug cocktail exerted significant neuroprotection in both experimental models thus demonstrating that simultaneous targeting of several pathophysiological pathways involved in the evolution of ischemic injury may represent a rational therapeutic strategy for stroke.