Exogenous administration of tumor necrosis factor-α has been shown to both enhance and attenuate cutaneous healing in a dose-dependent manner. We examined the effects of tumor necrosis factor inhibition in the healing wound by both systemic and local administration of tumor necrosis factor-binding protein. Male Balb/C mice underwent dorsal skin incision with subcutaneous implantation of 20 mg polyvinyl alcohol sponges (4 per animal). In Experiment I, one group (n= 20) received intraperitoneal injections of tumor necrosis factor-binding protein (3 mg/kg) at the time of wounding, while another group (n= 20) received saline. Four animals from each group were euthanized on days 1, 3, 5, 7, and 14 postwounding. In Experiment II, one group (n= 10) received an intraperitoneal injection of tumor necrosis factor-binding protein (3 mg/kg) at the time of wounding and every third day thereafter. Another group (n= 10) received an intraperitoneal injection of saline at the time of wounding and every third day thereafter. In Experiment III, one group received a single intraperitoneal injection of tumor necrosis factor-binding protein (3 mg/kg) at the time of wounding (n= 7), or on postwounding day 4 (n= 7), or day 7 (n= 7). Another group received saline injections at the time of wounding (n= 7), or on postwounding days 4 or 7 (n= 7, respectively). All animals in Experiments II and III were killed at postwounding day 14. Wound breaking strengths were assessed using a tensiometer. Wound fluid collected from the implanted sponges was assayed for tumor necrosis factor-α and tumor necrosis factor-binding protein levels using a biological assay and enzyme-linked immunosorbent assay, respectively. Collagen gene expression in sponge granulomata was assessed by Northern analysis. Collagen deposition in sponges was quantified by measuring hydroxyproline content. Wounds were significantly weaker in the animals that received repeated injections of tumor necrosis factor-binding protein with a mean wound breaking strength of 93.1 g vs. 186.6 g in controls (p < 0.05). Wound breaking strength in groups that received a single injection of tumor necrosis factor-binding protein on either day 0, 4, or 7 postwounding were no different than their respective controls. There was no difference in the mean hydroxyproline content of sponges between any of the tumor necrosis factor-binding protein groups and their respective controls. Northern analysis for collagen I and III expression also revealed no differences. These data indicate that continued systemic administration of tumor necrosis factor-binding protein resulted in significantly weaker wounds with no corresponding differences in wound collagen content, and collagen gene expression. This suggests that tumor necrosis factor-α inhibition throughout healing leads to a qualitatively impaired wound without a quantitative alteration in collagen deposition.