Objective. The aim of this prospective study was to identify latent phase predictors of active labor duration. Design. Prospective clinical study. Setting. Two delivery units in Sweden. Sample. Healthy nulliparous women with a normal pregnancy, spontaneous onset of active labor at term, and a cervical dilatation of 4 cm or more on admission to the delivery ward (n = 2,072). Methods. The women were asked to answer questions concerning their food and fluid intake, amount of rest and sleep during the preceding 24 hours and to assess their labor pain, sense of security and expectations of the childbirth on a visual analog scale (VAS). Duration and intervals of contractions, cervical dilatation, and position of the fetal head were noted by the midwife. A multiple regression analysis was performed with active labor duration as the outcome variable. Main outcome measure. Predictive factors of active labor duration. Results. Normal food intake during the preceding 24 hours was associated with short labor duration. A long latent phase, low levels of assessed labor pain and few hours of rest and sleep during the preceding 24 hours were significant independent predictors of extended active labor duration, when high birth weight, long contraction intervals, slight cervical dilatation, intact membranes within 2 hours of admission, high maternal age and malposition of the fetal head were controlled for. Conclusion. New findings are that latent phase duration as well as food intake and the amount of rest and sleep during the preceding 24 hours are independent predictors of labor duration.