Several criteria are suggested for measuring the effectiveness of separation systems (chromatography, electrophoresis, etc.) in one and two dimensions in the presence or absence of gradients. A major point is that resolution is truncated so that if two peaks are separated by more than some sufficient amount, ΔS, their effective separation remains constant. This eliminates overemphasis on large separations which tend to obscure less well-separated peaks. It is shown that the qualitative behavior of at least two of the suggested class of criteria are similar, and that the criteria are insensitive to arbitrary parameters over a wide range. Illustrative examples are given from one-and two-dimensional gel electrophoresis, including the effects of pore gradients and/or ionic detergents (e.g., sodium dodecyl sulfate).