A new study reveals some experimental antibody treatments for neurodegenerative disorders increase neuroinflammation. The effect was seen in human microglia, not mouse microglia. This may explain why mouse studies have shown these treatments to be effective, but the effect has not been replicated in human studies.