Natural antibodies are used to compare immune systems across taxa, to study wildlife disease ecology, and as selection markers in livestock breeding. These immunoglobulins are present prior to immune stimulation. They are described as having low antigen specificity or polyreactive binding and are measured by binding to self-antigens or novel exogenous proteins. Most studies use only one or two antigens to measure natural antibodies and ignore potential effects of antigen specificity in analyses. It remains unclear how different antigen-specific natural antibodies are related or how diversity among natural antibodies may affect analyses of these immunoglobulins. Using genetically distinct lines of chickens as a model system, we tested the hypotheses that (1) antigen-specific natural antibodies are independent of each other and (2) antigen specificity affects the comparison of natural antibodies among animals. We used blood cell agglutination and enzyme-linked immunosorbent assays to measure levels of natural antibodies binding to four antigens: (i) rabbit erythrocytes, (ii) keyhole limpet hemocyanin, (iii) phytohemagglutinin, or (iv) ovalbumin. We observed that levels of antigen specific natural antibodies were not correlated. There were significant differences in levels of natural antibodies among lines of chickens, indicating genetic variation for natural antibody production. However, line distinctions were not consistent among antigen specific natural antibodies. These data show that natural antibodies are a pool of relatively distinct immunoglobulins, and that antigen specificity may affect interpretation of natural antibody function and comparative immunology.