Preterm delivery is birth occurring before 37 completed weeks of gestation. Preterm birth is the primary cause of morbidity and mortality in children especially if this occurs before 34 weeks of gestation. If preterm birth could be predicted and treated accordingly, this would greatly reduce mortality, morbidity and associated costs. There have been many attempts to develop an accurate and efficient method to predict preterm premature rupture of membranes (PPROM) and preterm labor that leads to spontaneous preterm birth (SPB). However, the initial signs and symptoms are most often mild and may even occur in normal pregnancies, making early detection rather difficult. The aim of this paper is to provide an overview of the current methods employed in predicting preterm birth occurring due to preterm labor and PPROM. Among these methods are risk scoring systems, cervical/vaginal screening for fetal fibronectin, cervical assessment by ultrasonography, uterine activity monitoring, biomarkers such as endocrine factors, cytokines and enzymes, fetal DNA and genetic polymorphism. SPB is multifactorial, and so it is highly unlikely that a single test can accurately predict SPB. A combination of biological markers is also reviewed in the estimation of the risk of preterm delivery.