Criteria for the success of a CELSS type experiment must be relevant to the optimization of yield under conditions close to those expected in space. Additionally it is highly unlikely that the same resource will be limiting both in space and on planetary bodies. Therefor it will be necessary to develop two or more sets of criteria, for it is difficult to determine the optimal conditions for one resource when growth is limited to a fraction of its potential by an other resource. The development of criteria is further complicated by the high degree of interaction between parameters in biological systems. In most systems many nonlimiting parameters have a high degree of flexibility, but the optimum setting for these parameters will depend upon the resource being maximized. The data base for each parameter should cover its whole response surface in order to have a predictive value. Each CELSS system protocol should be designed to maximize the limiting resource and should be tested at the desired level of yield.