The research addressed the question: Does an ecologically-based model contribute to understanding the etiology of preterm birth? The study was a secondary analysis of data from two prospective studies of the relationship between genital tract colonization and shortened pregnancy at the University of Colorado Health Sciences Center from 1985 to 1988. The sample was four hundred and thirty-three essentially well women representative of the population of medically indigent clients seeking antepartum care in the Denver metropolitan area. The study identified a distinct structure of six genital microecology clusters at midpregnancy utilizing multivariate cluster analysis methodology. Hierarchical multivariate discriminant function analysis was applied to examine the ecological model for the ability to identify women who experienced preterm labor and birth. The host characteristics prior preterm birth, prior miscarriage and stress and the microecology attributes of cervical Chlamydia trachomatis colonization, Staphylococcus aureas colonization and treatment of vaginitis in the context of a cluster where bacterial vaginosis associated with Gardnerella vaginalis was present were associated with the discriminant function to identify women who experienced preterm birth. Recommendation for nursing practice in antepartum care include demonstration projects and clinical trials of Chlamydia trachomatis treatment, utilization of the most prepared nursing personnel in initial health history, epidemiologic assessment and care-planning as well as increased vigilance in recognizing the ecological context in the treatment of vaginitis. The study findings support a growing body of research evidence which suggests a microbially-mediated etiology of preterm birth which is potentially modifiable through antepartum care practices.