Speech segmentation is one of the first tasks infants face when learning their mother tongue. On the one hand, it has been argued that statistical learning could function as a gateway to speech segmentation in the absence of pre-existing knowledge about the language to be acquired. Initial support for this hypothesis came from corpus analyses that showed that Transitional Probabilities (TPs) tend to be higher within words than at word boundaries, and that infants can segment continuous streams of stimuli in various perceptual domains by calculating TPs (Saffran et al., 1996). On the other hand, infants could also segment speech with prosodic cues, such as lexical stress, phrasal prominence and intonational contours which they are sensitive to during the first months of life, if not already at birth (e.g., Johnson & Jusczyk, 2001). Here we present evidence about how infants weigh statistical and prosodic information when segmenting continuous speech. We argue that the dominant idea that infants discover their native language prosodic structures through statistical regularities (Thiessen & Saffran, 2003), as has been evidenced in English- and French-learning infants, is not found with German-learning infants. With more natural speech stimuli, infants only become sensitive to statistical regularities in the speech signal by their first birthday (Marimon et al., 2022; 2024).