The ability to differentially batch-mark several size-groups of fish stocked concurrently in lakes or rivers makes it possible to compare survival within a single year. We evaluated stocking of walleyes Stizostedion vitreum in Illinois during 1991–1996. Ten lakes were stocked with combinations of two or three sizes of walleyes, including fry and small (50-mm) and large (100-mm) fingerlings. This permitted a total of 73 lake-year comparisons: fry versus large fingerlings (23 lake-years), fry versus small fingerlings (29 lake-years), and small versus large fingerlings (21 lake-years). Stocked fish were differentially marked with oxytetracycline or fin clips. Electrofishing catch per effort, relative survival, and population estimates were used in conjunction with production costs to compare size-based contributions and survival. Survival generally favored fingerlings over fry (70% of lake-years for small fingerlings, 67% for large fingerlings) and small fingerlings over large fingerlings (72% of lake-years). In some lake-years, fry and large fingerlings had the highest survival. Fish stocked as fry and small fingerlings were larger than large fingerlings beginning their third year of growth. Based on our results, stocking small fingerlings will generally be more cost-effective than stocking fry or large fingerlings. However, if fry or large fingerlings are available, stocking assessment on specific lakes may allow biologists to determine particular lakes where alternative sizes could be economically feasible.