The objective was to predict interannual fluctuations in the size of sea-trout fry when they emerged from the redd, using models developed from field data for 70 excavated redds (≥three per year), and from experimental data on egg and alevin development at 30 constant temperatures in the laboratory (range 1·5—10·5) C with 100 naturally fertilized eggs at each temperature). Egg weight increased with female length and also with the number of eggs laid in a redd, both relationships being well described by a power function. Early spawners were the largest females laying the largest and most numerous eggs, whilst late spawners were the smallest females laying the smallest and least numerous eggs, with middle spawners being intermediate between these two extremes. Mean values for egg weight and number of eggs per redd were obtained for these three groups. The numbers of early, middle and late spawners for each year of a 30-year study and the mean values from the excavated redds were used to estimate weighted means for the number of eggs per unit area and egg weight. Mean values varied considerably between years (30-year ranges: 518–7964 eggs per 60 m2; 112–138 mg wet weight). In the laboratory, mean weights of newly hatched alevins and newly emerged fry were both related positively to mean egg weights. Alevin and fry mean weights were independent of the number of days required for 50% of the eggs to hatch or fry to emerge. Models described in a previous paper formed the basis of those used to predict fry weights over the emergence period. Model predictions were validated by field data for the whole emergence period in 8 years (1967–1971, 1974, 1975, 1980), and by pre-fry weights on single dates in 21 years (1967–1987). As pre-fry densities on these single dates were very similar to egg densities for the same year class, mortality in the egg and alevin stages was very low. The chief objective was therefore fulfilled, and the extent of interannual fluctuations for the 30-year study showed some variation in mean fry weight (30-year ranges: 153–193 mg for both the whole emergence period and the date on which 50% of fry emerged) but a progressive decrease in fry weight through the emergence period. Possible reasons for this variation are discussed, and it is concluded that the size of the female spawners is the dominant factor.