A method of determining the rate of natural mortality, based on the construction of a regression between the rate of annual fishing mortality (determined by VPA) and fishing effort is described. The intersect of this regression must be equal to zero if the natural mortality rate has been correctly identified and something other than zero if it has not been correctly chosen. Data for Champsocephalus gunnari in Subarea 48.3 are used to show that the described method is sound. When a mistake is made in the choice of natural mortality coefficient, the estimate of the intercept of the regression is other than zero. A bias in the intercept will be positive if the assumed natural mortality coefficient is less than the “true” value and close to zero or slightly greater than zero if the assumed value is greater than the “true” one.