BACKGROUND: We sought to develop a model of bacterially induced preterm delivery in rats to parallel similar models in mice. METHODS: Female Sprague-Dawley rats on day 17 of gestation (normal term = 21-22 days) were inoculated into the uterus with either 2 x 10(9)-7 x 10(10) killed E. coli organisms, 1-4 x 10(8) live E. coli or sterile solution. These inoculations were made either via trans-cervical catheter or by direct intrauterine injection at laparotomy. Animals were then observed for delivery for variable periods up to term. Necropsies were performed and fetal viability was assessed. RESULTS: No rats delivered prematurely after bacterial exposure (27 animals observed for at least 48 hours), and all animals followed to term (n = 3) delivered live pups. No dams exhibited signs of systemic illness. There was a statistically significant but small negative effect of killed E. coli on fetal viability (100% of 80 fetuses from 6 control pregnancies and 93% of 182 fetuses from 14 bacterially-treated pregnancies were alive at necropsy, p = 0.014). Live bacteria had a larger effect on fetal viability, with only 64% of 14 fetuses, 47% of 28 fetuses and 32% of 31 fetuses surviving after trans-cervical administration of 7 x 10(7), 2 x 10(8) and 4 x 10(8) E. coli, respectively. CONCLUSION: Unlike mice, it has proven difficult to induce preterm labor in the rat using E. coli as a stimulating agent. The relevant literature is reviewed and hypotheses are offered to explain this phenomenon.