Glial cells not only serve supportive and nutritive roles for neurons, but also respond to protracted stress and insults by up-regulating inflammatory processes. The complexity of studying glial activation in vivo has led to the widespread adoption of in vitro approaches, for example the use of the bacterial toxin lipopolysaccharide (LPS, a ligand for toll-like receptor 4 (TLR4)) as an experimental model of glial activation. Astrocyte cultures frequently contain minor numbers of microglia, which can complicate interpretation of responses. In the present study, enriched (≤5% microglia) astrocytes cultured from neonatal rat cortex and spinal cord were treated with the lysosomotropic agent L-leucyl-L-leucine methyl ester to eliminate residual microglia, as confirmed by loss of microglia-specific marker genes. L-Leucyl-L-leucine methyl ester treatment led to a loss of LPS responsiveness, in terms of nitric oxide and cytokine gene up-regulation and mediator (pro-inflammatory cytokines, nitric oxide) output into the culture medium. Surprisingly, when astrocyte/microglia co-cultures were then reconstituted by adding defined numbers of purified microglia to microgliadepleted astrocytes, the LPS-induced up-regulation of pro-inflammatory gene and mediator output far exceeded that observed from cultures containing the same numbers of microglia only. Similar behaviors were found when examining interleukin-1β release caused by activation of the purinergic P2X7 receptor. Given that astrocytes greatly outnumber microglia in the central nervous system, these data suggest that a similar interaction between microglia and astrocytes in vivo may be an important element in the evolution of an inflammatory pathology.