Objective: The purpose of this study was to examine the roles of intrauterine infection, inflammation, and spontaneous labor on the expression of matrix metalloproteinase-9 in fetal membranes and amniotic fluid of late pregnant rhesus monkeys. Study Design: Pregnant rhesus monkeys with timed gestations were chronically catheterized to allow serial sampling of amniotic fluid before and during experimentally induced intrauterine inflammation. Six animals received group B streptococci into the chorionic-decidual space, and 4 animals received intra-amniotic interleukin-1β infusions (10 μg). Three additional animals were serially sampled by amniocentesis through late pregnancy until spontaneous term labor. Amniotic fluid samples were examined by zymography for matrix metalloproteinase-9 and -2 and Western immunoblot for matrix metalloproteinase-9 and -2 and tissue inhibitors of metalloproteinase-1 and -2. Fetal membranes were obtained at cesarean delivery during labor (before rupture), formalin fixed, and embedded in paraffin for immunocytochemistry of matrix metalloproteinase-9 and in situ hybridization of matrix metalloproteinase-9 messenger RNA. Tissues from 2 additional animals were collected as age-matched nonlabor controls. Results: In amniotic fluid, the 92-kd latent matrix metalloproteinase-9 was detectable in late pregnancy and increased dramatically, followed by the appearance of the 83-kd active form before spontaneous term delivery. Amniotic fluid matrix metalloproteinase-2 levels (both latent and active forms) remained relatively constant throughout pregnancy and in labor. Both bacteria and interleukin-1β rapidly increased the signal of latent matrix metalloproteinase-9 without a consistent increase in the active form before the onset of labor. Chorionic-decidual inoculation of group B streptococci was followed by the expression of latent matrix metalloproteinase-9 before the appearance of group B streptococci in amniotic fluid or the onset of labor. Matrix metalloproteinase-2 increased to a new steady-state level or remained unchanged after group B streptococci inoculation or interleukin-1β infusion, respectively. Amniotic fluid tissue inhibitors of metalloproteinase-1 declined and tissue inhibitors of metalloproteinase-2 remained unchanged during early group B streptococci infection, after interleukin-1β infusion and on the day of spontaneous term labor. However, both tissue inhibitors of metalloproteinase-1 and -2 levels increased after preterm labor that was induced by group B streptococci. Immunocytochemistry localized matrix metalloproteinase-9 protein to amnion and chorion epithelial and mesenchymal cells and decidual stromal cells. Granular matrix metalloproteinase-9 staining was observed in the connective tissue layer of inflamed fetal membranes. In situ hybridization for messenger RNA confirmed the production of matrix metalloproteinase-9 by amnion and chorion. Conclusion: Bacterial- and interleukin-1β–induced preterm labor and spontaneous term labor are preceded and accompanied by progressive increases in amniotic fluid matrix metalloproteinase-9 (92 kd) in rhesus monkeys. Amniotic fluid matrix metalloproteinase-9 may serve as a clinical marker for the onset of both preterm and term labor. (Am J Obstet Gynecol 2002;186:128–38.)