The dace (Leuciscus leuciscus (L.)) is a small, shoaling, rheophilic, lithophilic cyprinid fish, and is an important component of chalk stream ecosystems. Historical recaptures of marked dace in the River Frome suggested that some individuals were mobile, however it was unclear whether these movements occurred predictably in time or space. Migrations, defined as deliberate, temporally predictable translocations in space, allow individuals to exploit new resources, avoid predators and find a mate. However, energy is generally utilised during locomotion, and there are potential costs associated with moving to an unfamiliar environment, including reduced foraging opportunities, and increased predation risk. To be successful, individuals must find food, avoid predation and reproduce. If adult dace can solve these problems efficiently in one section of river, then considering the potential costs associated with migration, it would be most profitable for them to remain in this area. A range of techniques were used to study the migrations and habitat use of adult dace in the River Frome, covering both daily and seasonal cycles. The movements of 32 radio-tagged dace were tracked over a period covering over 16,000 hours. During this period 6864 radio locations were made, and radiotagged fish covered a combined minimum distance of over 109 km. A further 1724 dace were individually or batch marked, of which 128 were recaptured. In addition over 4000 dace were recorded on video as they passed through a fish counter. The daylight habitat use of 377 adult dace was visually observed from the bank, and habitat suitability indices created. The nocturnal habitat use of 11 adult dace was assessed using radio-active isotope tags, which were located on 247 occasions over a period covering 121 "fish nights". The adult dace observed in this study moved extensively, and their movements were both temporally and spatially predictable, and therefore constituted migrations. Of the 30 fish released at their capture sites, at least 19 moved more than 500m fi-om this site during the observation period. Spawning occurred in millstreams, and was preceded by an upstream migration. After spawning, slow flowing shaded sites out of the main river were selected by at least five out of six radio-tagged dace. A substantial upstream migration was observed in late spring, and diel migrations between separate distinct day and night habitats were observed during the summer. In autumn adult dace moved downstream and formed aggregations in the tidal reaches. Consequently, the hypothesis that dace occupy one section of the river throughout the year was rejected.