Marginal otolith increment width analysis was performed on field-collected larval and juvenile spring-spawned herring Clupea harengus that experienced variable feeding conditions and high temperatures that were above the optimum for growth. Although drastic zooplankton biomass reduction had a significant effect on increment width, a delay of a few days in the otolith response was observed. More importantly, a very clear, positive temperature effect on marginal increment width was demonstrated in fish characterized by temperature independent somatic growth. These results indicate that under natural conditions it may be impossible to distinguish increment width changes related to variation in feeding conditions from changes caused by temperature fluctuations. Therefore, it was concluded that marginal otolith increment width analysis could not be used as a recent growth index (IG) for herring larvae and juveniles exposed to drastic temperature fluctuations. The implication of these results is significant not only for the use of marginal increments as a recent growth index, but also if growth rate backcalculation is to be used as a research method.