Preterm birth (<37 weeks) complicates 12.5% of all deliveries in the USA, and remains the leading cause of perinatal mortality and morbidity, accounting for as many as 75% of perinatal deaths. Despite the recent temporal increase in preterm birth, efforts to understand the problem of prematurity have met with little success. This may be attributable to the under-appreciation of the etiologic heterogeneity of preterm birth as well as the heterogeneity in its underlying clinical presentations--spontaneous onset of labor, preterm premature rupture of membranes, and medically indicated preterm birth. In this paper, we review data regarding preterm births with particular focus on its incidence, temporal trends, and recurrence. Studies of births from the USA indicate that the recent temporal increase in the overall preterm birth rate is driven by an impressive concomitant increase in medically indicated preterm birth. However, the largest temporal decline in perinatal mortality has also occurred among medically indicated preterm births (relative to other clinical subtypes), suggesting that these obstetric interventions at preterm gestational ages are associated with a reduction in perinatal mortality. Recent data indicate that spontaneous preterm birth is not only associated with increased recurrence of spontaneous, but also medically indicated, preterm birth, and vice versa. This suggests that the clinical subtypes may share common underlying etiologies. Since medically indicated preterm birth accounts for as many as 40% of all preterm births, efforts to understand the reasons for such interventions and their impact on short- and long-term morbidity in newborns is compelling. Further research is necessary in order to understand the mechanisms and etiology of preterm birth, thus leading to the possibility of effective preventive or therapeutic strategies.