1. Marine life in offshore regions has not been fully censused, yet related conservation policy relies on our ability to identify areas of high biodiversity. 2. We assessed the census of marine finfish on the Scotian Shelf, Northwest Atlantic using data collected during annual research vessel surveys between 1970 and 2000. The species accumulation curve did not reach an asymptote reflecting that new species continued to be discovered throughout the survey period. Only 0.12% of the area of the Scotian Shelf has been sampled since 1970. 3. Since 1974, when over 50% of the species had been discovered, the community composition has been relatively constant. However, the dominance structure has changed dramatically as reflected in the geographic contraction of the formerly abundant, large-bodied piscivores concomitant with the geographic expansion of their prey species. 4. The region is under-sampled, and species' distribution and abundance are changing. A precise estimate of diversity is elusive. As an alternative, we searched for physical correlates of finfish diversity to identify its possible surrogates. Surrogates have potential both as a method for understanding process and as a tool for conservation management. We examined the effect of area and depth range on species richness. High species richness was associated with larger areas and greater depth range at large spatial scales. 5. Highly diverse areas include the Bay of Fundy, the Eastern Gully, the slopes, Western Bank and the northeastern shelf. Until now, the northeastern shelf has been under-appreciated as a highly diverse area. Such information will be important for environmental impact assessments as well as selection of 'sensitive' or protected areas.