The hybridoma technique has been shown to be a most reproducible method for producing rodent monoclonal antibodies but poor results were obtained when it was used for generating human monoclonal antibodies. For immunotherapy, murine monoclonal antibodies are inadequate, whereas human monoclonal antibodies are virtually indispensable. Cellular, chemical, genetic and molecular methods to generate human monoclonal antibodies have been developed. Most often, the monoclonal antibodies for therapy are selected after deliberate vaccination, according to their high affinity towards an arbitrarily-chosen epitope of a pathogen or cellular antigen and therefore the selection is obviously skewed. A major hindrance of the production of therapeutic human monoclonal antibodies is the lack of an appropriate strategy to define and select the antibodies that would be effective in vivo. In contrast to antibodies induced by vaccination, there has been only a marginal interest in monoclonal antibodies which reflect antibodies of the innate immunity. In the future, human monoclonal antibodies that resemble antibodies that are ubiquitously present in sera of healthy individuals might serve as novel therapies in diseases such as Alzheimer's disease, where no other therapy exists.