The dispersal of fish larvae during the early life stages plays an important role in recruitment. Together with oceanographic processes, larval orientation and swimming behavior significantly influences dispersal. However, currently there is no information of larval behavior in situ for most subpolar species. The Atlantic herring (Clupea harengus) is an ecologically and commercially important component of the North Atlantic ecosystem. This species has sustained a large fishery for over a century, and its stocks have experienced both dramatic collapse and recovery. The Atlantic herring is highly migratory, making it challenging for researchers to determine with certainty the structure and discreteness of different herring stocks. In this context, studying the behavior of herring larvae at sea is crucial for a deeper understanding of their dispersal. However, to date information on the orientation behavior and swimming abilities of herring larvae in situ is missing. In this study, we coupled in situ observations and laboratory experiments to investigate whether herring larvae display orientation when swimming in situ, and which mechanisms/cues they could use. We video recorded the orientation behavior of 208 herring larvae between 14 and 28 days post hatch (DPH) in the coastal Norwegian North Sea while they drifted in transparent behavioral chambers. We also tested the orientation of 136 larvae in a magnetic laboratory, in which they were deprived of any external cue and where we could modify the direction of the magnetic field. We report evidence that herring larvae have a significant preferred orientation direction to the southeast, which does not change between 14 and 28 DPH. Moreover, our results suggest that sunlight plays a key role as larvae had a highly significant orientation towards the sun during sunny weather, but they lost this ability, and exhibited lower precision in their orientation, under an overcast sky. We did not find evidence of magnetic compass orientation, indicating that the orientation direction of herring larvae is not magnetic, at least at this life history stage. Larvae swam at an average speed of 0.36–0.40 cm/s and reached maximum speeds of 3–3.36 cm/s. These results demonstrate that 14–28 DPH herring larvae are capable of orienting in situ. Possible implications of this orientation behavior for larval transport in Norwegian waters are also explored.