Teleosts such as tunas and billfish lay millions of tiny eggs weighing on the order of 0.001 g, whereas chondrichthyes such as sharks and rays produce a few eggs or live offspring weighing about 2% of adult body mass, as much as 10 000 g in some species. Why are the strategies so extreme, and why are intermediate ones absent? Building on previous work, we show quantitatively how offspring size reflects the relationship between growth and death rates. We construct fitness contours as functions of offspring size and number, and show how these can be derived from juvenile growth and survivorship curves. Convex contours, corresponding to Pearl Type 1 and 2 survivorship curves, select for extremes, either miniscule or large offspring; concave contours select for offspring of intermediate size. Of particular interest are what we call critical straight-line fitness contours, corresponding to log-linear Pearl Type 3 survivorship curves, which separate regimes that select for opposite optimal offspring sizes.