The diel vertical migration (DVM) and population dynamics of northern krill were investigated in the Clyde Sea over several years using depth-discrete net-sampling and a moored 300 kHz acoustic Doppler current profiler. Krill performed DVM throughout each year of sampling. The respective arrival and departure of the krill scattering layer to and from the upper layers occurred consistently 30 min after sunset and 30 min before sunrise. DVM of males and females was different, with females migrating closer to the surface at night than males of equivalent size. The sex ratio was initially 1:1 when juveniles matured into adults after their first winter, but the ratio became increasingly biased towards males as the season progressed such that, by October, the ratio was close to 3:1. A deterministic individual-based model, in which predation risk was a function of the light available to visual predators, showed that the bias could be accounted for by the difference in risk taken by males and females as a result of their DVMs. The same model also showed that, during summer, the difference in these DVMs resulted in females making a net energy gain that was 40% higher than that of males. The predicted net energy gain for females met the immediate energetic demand for egg production as well as providing a surplus that could offset any future decline in food availability. The model did not predict the observation that the decline in male numbers was greater than that of females over the winter. This may be a result of females being more able to cope with starvation through reabsorption of their lipid-rich ovaries. The greater demand for energy to fuel reproduction appeared to be driving females to undertake a riskier DVM than males. This drive is likely to be common to most euphausiid species.