A study was made of the reproductive biology and ecology of the young stages of perch in two shallow lakes on Deeside, Scotland, from 1975 to 1977. Emphasis was placed on quantifying production and mortality of embryonic, larval, juvenile and adult periods in the life history and in identifying factors limiting the success of these various stages. The contribution of the young stages of perch to the freshwater ecosystem and the relationship to other trophic levels, including potential predators and the zooplankton community, was assessed. Development of the ovary was described by histological analysis and related to changes in ovary weight expressed as a gonado-somatic index. There was no evidence that pre-ovulatory degeneration of oocytes occurred and residual oocytes after spawning were rare, thus population fecundity was not affected. The embryonic phase was characterised by a total mortality of only 4.7% which is attributed to the unique egg membrane which is thick and mucilaginous. Mortality during the pelagic larval phase was 4.3% per day which declined to 2% when fry attained the demersal habit, a mortality curve consistent with Marr's description of a high initial decline in abundance which levelled off with time. Despite variations in larval abundance between the years and lakes/mortality rate did not vary and the pelagic larval phase cannot be held responsible for any variations in year class strength. Food supply was assessed and suitable zooplankters found to be most frequent on larval hatch and sufficiently abundant for maintenance, although not always for growth near the end of the pelagic stage. Extremely high mortality was identified at Davan in 1976 between larval hatch and recruitment to the pelagic stock. It is suggested that the exposed and shallow nature of the spawning areas at Davan resulted in mechanical damage to larvae and young fish may have been driven into reedbeds inshore. Fluctuations occurred in year class strength and, although the strength of certain year classes was synchronous in Kinord and Davan, lack of synchrony in other years suggested that climatic conditions may not always be a major controlling factor. No correlation could be found between year class strength and air temperatures in excess of 10°C. Predation by adult pike and perch, which is particularly high on O+ fish during the first summer, may be an important regulatory factor and may explain the low abundance of adult perch III+ (0.20 - 0.39 kg wet weight/ha) and the variation in strengths of year classes. Annual production in the pelagic, larval phase was therefore important, being six to seven times as high as that of adult perch at Kinord. Grazing by pelagic larvae had only a slight effect on total zooplankton standing crop in late spring. However, predation of certain Cladocera was intense and limited abundance on a number of occasions. Predation by perch larvae enhanced declines initiated by other agents in the Cyclops population at Davan in 1976 and Diaptomus stocks at Kinord in both years. A marked decline in abundance of Cyclops at this loch in 1977 was coincident with high losses attributed to removal by perch larvae and a reduction in the Cyclops population at Davan in 1977 was synchronous, with predation on predominantly younger copepodite instars.