An understanding of the factors controlling the population level of marine fish is one of the most important objectives of marine biologists because of its bearing on the productivity of the sea in products useful to man, and, it might also be added, because we know so little about it. This paper will discuss some of the developments in a limited part of this field applying to haddock, one of the most important species of fish found off the New England Coast. Although this fish population is limited in comparison with the oceans as a whole, it is extensive in comparison with many terrestrial populations. In numbers it lies in the hundreds of millions; in extent it ranges over an area greater than the combined New England States without Maine and Vermont; while in observability it might be compared to a terrestrial population of comparable size shrouded in a dense fog when viewed from a dirigible cruising at a height of a few hundred feet (Fig. 1). The first problem is to identify and evaluate the principal factors which control the population level of this species. Since it is impossible or impractical to control the environmental conditions, our procedure is to measure certain of the important environmental factors and compare their variations with variations in the population level. Probably the greatest present limitation to progress in this field is the accuracy with which we can measure population size and population changes. If you question this statement, just consider how far beyond the speculative stage, knowledge of temperature, salinity, food supply, and other environmental factors will take you by itself. However, if you can obtain accurate data showing the changes in the population level which accompany variations in environmental conditions, you can begin to find out which factors are dominant and at what level they become important. To measure population size, the most productive method now available to the biologist is to obtain records of fishing effort and catch, either commercial or sport, and from these calculate the returns perunit-of-fishing-effort. This generally is considered as representative of the average density of the fish and proportional to the total population in the sampled area. Unfortunately, this is only approximately true, AS