Treatment of Alzheimer disease or amyotrophic lateral sclerosis with anti-inflammatory drugs (to prevent disease or slow its progression) has yielded mixed results, despite evidence indicating that local cytotoxic inflammation occurs in these conditions. Here, through consideration of the importance of immune cell origin (resident versus blood-derived immune cells) and activity (pro-inflammatory versus anti-inflammatory activity) under neurodegenerative conditions, we propose a model that reconciles these seemingly inconsistent data. We suggest that systemic immune cells (CD4⁺ T cells and peripheral blood-derived monocytes) must be recruited to the CNS to modify potentially destructive local inflammation, and that the failure of systemic anti-inflammatory drug therapies to arrest neurodegenerative disease progression might result from drug-induced suppression of such recruitment. Thus, we propose that an appreciation of the distinctive temporal and spatial contributions of resident and systemic leukocytes to disease progression is essential for the development of effective therapeutic regimens.