Previous data support the hypothesis that during inflammation, interleukin (IL)-1 beta and IL-6 are involved in fever, in activation of the hypothalamic-pituitary-adrenal (HPA) axis, and in the induction of eicosanoids. Most of the pathophysiologic effects of IL-1 beta and Il-6 are mediated by prostaglandins (PGs), modulated by other cytokines, and antagonized by glucocorticoids (GC), a final product of the HPA axis. To further test these relationships, we measured changes in body temperature using biotelemetry in mice deficient in genes for IL-1 beta and/or IL-6 (IL-1 beta knockout [KO] and IL-6 KO) following injection with lipopolysaccharide (LPS) to induce systemic inflammation or turpentine to induce local abscess. Circulating IL-6, tumor necrosis factor alpha (TNF-alpha), GC, and PGE2 were measured in these mice after treatment. IL-1 beta KO mice responded with reduced fever and IL-6 KO mice with normal fever to a high dose of LPS. In contrast, neither type of KO mice produced fever to turpentine. PGE2 levels (measured in the circulation) were suppressed in both types of KO mice injected with turpentine. IL-1 beta KO mice showed deficiency in IL-6 following turpentine, but not LPS, injection. LPS-induced increases in TNF-alpha did not differ between IL-1 beta KO mice and their wild-type counterparts, whereas IL-6 KO mice showed exacerbated LPS-induced circulating TNF-alpha. No differences were noted in plasma elevations of GC between KO and wild-type mice following injection of LPS or turpentine, indicating that IL-1 beta and IL-6 are not required for activation of the HPA axis during inflammation. Our data demonstrate that in the mouse, IL-1 beta and IL-6 are critical for the induction of fever during local inflammation, whereas in systemic inflammation they appear only to contribute to fever.