Catalytic antibodies have been developed by experimental approaches exploiting the analogy between antibody-antigen and enzyme-substrate interaction. Haptens have been prepared to model the electrostatic or geometric attributes of a reaction's transition state and to induce combining sites having appropriate catalytic residues. The relative merits of these design strategies may be gleaned from the apparent activities and efficiencies of the respective catalysts. The implications of screening strategies on the kinetic characteristics of the resulting abzymes are also considered. Combining-site hypermutation provides the variation in the antibody repertoire from which high-affinity clones are selected. The same mechanism can also lead to a subset of antibodies with reduced hapten affinity, but improved catalytic activity. This possibility has not been adequately characterized, but is suggested by a number of considerations. These include the unexplained efficiency and diversity of mechanisms utilized by various antibody catalysts, and the observed catalytic activity of antibodies found in autoimmune serum. This article attempts to assess critically the evidence for rational design of catalytic activity in antibodies. Correlations among abzymes and their relevant models could lead to revised or novel strategies for producing better catalysts.