Variations in the production of carp, Cyprinus carpio, were measured in four consecutive seasons in nine similar, contiguous ponds. These were believed to be the most extensive uniformity studies made of fish production in ponds. The fertility of the carp ponds varied from year to year partly due to fertilization; the means and coefficients of variation of net production in kg/ha in replicate ponds within years were, respectively: 285, 0.179; 196, 0.143; 140, 0.227; 427, 0.123. Variations in weight increments made by similar numbers of tilapia, Tilapia mossambica, were measured over 50-day periods in 10-foot diameter plastic pools having meticulously standardized soil substrates. These data suggest: 1) that carrying capacity in ponds is less stable than commonly believed, 2) that no environmental factor, such as basin fertility, maintained a dominant, continuous influence, and 3) that in each new season production was controlled by a new assortment or combination of factors.