The diagnosis of spontaneous preterm birth (sPTB) remains a major obstetric challenge. Prediction is currently confined to evaluating historical risk factors, clinically assessing uterine activity and cervical changes, and measuring cervical length (CL) and fetal fibronectin (fFN). These approaches have not managed to inform treatment to improve neonatal outcome or to reduce sPTB incidence. Hence, there remains a need to develop new minimally-invasive and cost-effective techniques which can better risk stratify and predict sPTB. I tested whether a series of innovative spectroscopic and light-based technologies could hold predictive value for sPTB. I first employed Raman spectroscopy to assess mid-cervical cervicovaginal fluid (CVF) in pregnant women as a predictive tool for sPTB, and identified spectral changes in protein conformation, aminoacids, lactic acid, urea and acetate between birth outcomes. I then used Polarisation-sensitive optical coherence tomography (PS-OCT) on the in vitro cervix to assess collagen orientation which is thought to change preceding sPTB. PS-OCT distinguished the preferential orientation of different cervical regions, which was consistent with histology and previous X-ray diffraction findings. PS-OCT may prove useful to assess in vivo cervical remodelling pending the development of a probe. The cervical microstructure was further assessed with Magnetic induction spectroscopy (MIS). In vitro cervical transresistance significantly differed between the epithelial and stromal surfaces. When tested on pregnant women in a pilot preliminary study (the ECCLIPPx II study), MIS was shown to perform better than fFN and CL both in asymptomatic high-risk (AHR) and symptomatic (SYMP) women. Finally, metabolite concentrations in CVF were assessed with enzyme-based spectrophotometry within the ECCLIPPx II study. Combining CVF lactate, urea and acetate significantly improved the clinical performance of fFN and CL within the AHR and the SYMP cohorts. Overall, these new techniques hold promise for predicting sPTB, and support the need for larger fully-powered studies.