Background: Preterm delivery can be associated with substantial perinatal morbidity and mortality. Nearly in 5 to 13% of pregnancies, happening deliveries are preterm before 37 weeks' gestation. Thus, prediction of preterm labor in parturient can provide a window of opportunity to prevent these complications and to be better prepared to deal with them. Hence the current study has been conducted with an objective to assess the efficiency of cervical length and funneling as assessed by transvaginal ultrasonography in predicting preterm labor.Methods: The current study was a prospective study, conducted in the Department of Obstetrics Emergencies, Dhanalakshmi Srinivasan Medical College and Hospital between November 2015 to April 2017. The study population included 60 women with singleton pregnancy of gestational age between 28 + 0 and 36 + 6 weeks and with painful and regular contractions (>1/10 min for at least 1 hour). Cervical length, funnelling length measured at presentation and after 48 hours were considered as predictor variables. The outcomes of interest were delivery within 1 week, delivery before 37 weeks and delivery before 34 weeks.Results: There was a statistically significant difference in cervical length and funnelling length between term and preterm groups at the time of diagnosis and after 48 hours (P <0.001). The positive predictive value increased from 33.3% to 100% with cervical length ≤ 40 mm to ≤ 20 mm. The negative predictive value decreased from 100% to 79.2% with cervical length ≤40 mm to ≤20 mm. To predict a preterm delivery in threatened preterm delivery patients, the sensitivity of a cervical length of ≤25 mm was 77.7%, specificity was 95.2%, the positive predictive value was 87.5% and negative predictive value is 90.9%.Conclusions: Cervical length and funnelling as assessed by transvaginal ultrasonography are efficient in predicting preterm labor.