Following general life history theory, immediate reproductive investment (egg mass × fecundity/body mass) in oviparous teleosts is a consequence of both present and past environmental influences. This clarification questions the frequent use of season-independent (general) fecundity formulas in marine fish recruitment studies based on body metrics only. Here we test the underlying assumption of no lag effect on gametogenesis in the planktivorous, determinate-fecundity Atlantic herring (Clupea harengus) displaying large plasticity in egg mass and fecundity, examining Norwegian summer-autumn spawning herring (NASH), North Sea autumn-spawning herring (NSAH), and Norwegian spring-spawning herring (NSSH). No prior reproductive information existed for NASH. Compared with the 1960s, recent reproductive investment had dropped markedly, especially for NSAH, likely reflecting long-term changes in zooplankton biography and productivity. As egg mass was characteristically small for autumn spawners, although large for spring spawners (cf. different larval feeding conditions), fecundity was the most dynamic factor within reproductive investment. For the data-rich NSSH, we showed evidence that transient, major declines in zooplankton abundance resulted in low fecundity over several subsequent seasons, even if Fulton's condition factor (K) turned high. Temporal trends in K slope (K on total length) were, however, informative. These results clarify that fecundity is defined by (i) dynamics of primary (standing stock) oocytes and (ii) down-regulation of secondary oocytes, both processes intimately linked to environmental conditions but operating at different timescales. Thus, general fecundity formulas typically understate interannual variability in actual fecundity. We therefore argue for the use of segmented fecundity formulas linked to dedicated monitoring programs.fishery | fecundity | zooplankton | lag effect | recruitment