Data from early fisheries in the Northwest Atlantic and elsewhere suggest that catch rates of demersal species were very high despite primitive fishing methods. Circumstantial evidence suggests that earlier stocks may have been an order of magnitude greater than stocks in the last half-century. What was the structure of the ecosystem that supported these stocks? Relative to current ecosystem structure, alternative patterns involve very slow growth rates of the demersal species, very small pelagic stocks, negligible invertebrate predators, and efficient transfer of primary production to fish (no "detritus"). Each of these patterns suggests distinctive dynamics in pristine ecosystems, with different implications for the effects of fishing.