Both route and severity of infection may influence immunomodulator agents in sepsis. We studied the effect of each variable on HRL-3, an L-selectin-directed MAb that inhibits neutrophil function, in a rat sepsis model. Animals (n = 800) were randomized to be treated with either HRL-3 or placebo and to receive Escherichia coli either intravenously (IV) or intrabronchially (IB) in doses producing low or high mortality rates. Animals received antibiotics and were observed for 168 h. Route but not dose of E. coli altered the effects HRL-3 on mortality rate (mean hazards ratio +/- SE). With IV E. coli, compared with control, HRL-3 was beneficial and reduced the hazards ratio both early (0 to 6 h; -0.75 +/- 0.23) and late (6 to 168 h; -0.72 +/- 0.36) (P = 0.001 and 0.04, respectively, over all E. coli doses). In contrast, with IB E. coli HRL-3 reduced the hazards ratio early (-1.1 +/- 0.36) but worsened it late (0.87 +/- 0.23) (P = 0.002 for both effects over all E. coli doses) in patterns significantly different from IV E. coli (P < 0.0001). Compared with control, although HRL-3 did not alter lung neutrophil numbers or injury score at 6 or 168 h with IV E. coli (P = ns for all), it reduced both early and increased them late with IB E. coli (P </= 0.05 for all comparing 6 with 168 h). Thus immunomodulators inhibiting neutrophil function, although potentially beneficial with sepsis due to intravascular infection, may be harmful with extravascular infection regardless of severity.