We studied year-class strength (ycs) and length of age-0 Eurasian perch (Perca fluviatilis) in the top-down manipulated Bautzen reservoir, Germany, over a period of four years. In addition, density of main predators (perch and zander, age groups 1 to 3), water temperature and zooplankton biomass were recorded to determine whether these factors may have influenced yes. During the four years, yes varied substantially. Higher yes correlated with greater fish length in June and warmer water temperatures in spring, but there was no relationship with zooplankton biomass during that period. By contrast, the length of fish at the end of the year was negatively correlated to ycs, thus reflecting density-dependent growth, possibly because of the midsummer decline of Daphnia galeata. Variability in abundance of predatory perch and zander did not directly explain the year-class strength of age-0 perch. However, we suggest that length-dependent and hence temperature-dependent predatory mortality of age-0 perch in spring, may cause the observed variability in yes. Whereas temperature effects cannot be controlled by biomanipulation, the positive effect of high zooplankton biomass on perch growth, yes and early piscivory shows the central importance of perch in biomanipulation experiments.