ABSTRACT Otolith increment width and larval fish data (length and weight) were used to develop an individual-based model (IBM) to describe daily resolved growth rates of North Sea herring (Clupea harengus) larvae (Autumn Spawners) caught during International Herring Larvae Surveys in the ICES area IVa from 1990 to 1998. The model combines sagittal otolith readings (core and individual increment measurements), larval standard length and weight data, and solves an over-determined set of linear system equations for all parameters using the method of least square residuals. The model consists of a matrix, which describes the increment width formation of 119 larvae, a vector containing their length/weight measurements, and a vector describing residuals. The solution vector yields age-dependent maximum somatic growth rates of herring larvae up to an age of 41 days with sizes ranging from 10 to 25 mm. The observed environmental temperature in which larvae dwelled was relatively uniform. Therefore, measured increment width was individually used to determine daily growth from any single larva in relation to their potential maximum growth under optimal feeding conditions. The results are discussed with respect to the spatial and temporal variability of larval occurrence. Finally, an analysis of error estimation of the larval growth characteristics is presented.