Four "postulates", published by the bacteriologist and physician Robert Koch in 1890, form the basis of establishing the causative organism (i.e., pathogen) of an infectious disease. Koch received the 1905 Nobel Prize in Physiology or Medicine for his contributions, including the isolation of the cause of tuberculosis, Mycobacterium tuberculosis, and this theoretical work.
The principles, as originally stated, need elaboration and clarification, but remain some of the most important ideas in modern medicine.
While translations vary, a good set is:
- infected tissue must show the presence of a particular microorganism not found in healthy animals
- the microorganism must be isolated and grown in a pure culture
- when injected into a healthy animal, the microorganism must cause the disease associated with it
- this “second generation” microorganism should then be isolated and shown to be identical with the microorganism found in 1. This postulate does not appear in Koch's original presentation, but has generally become accepted.
While the original postulates work for many diseases, for a variety of reasons, Koch's original postulates may not be demonstrable. Later workers have proposed variants that reflect new capabilities in detecting or growing organisms, recognition of dormant states of infection, or the need for cofactors to trigger active disease.
Although there are many nuances, the current equivalent of the Postulates works largely when "nucleic acid sequence" can replace "organism".  There must be reasonable demonstration that the sequence is associated with the organism.
Several decades after Koch, but still before the explosive growth of techniques in infectious disease microbiology, Thomas Rivers, a virologist and president of the Society of American Bacteriologists, said, "It is unfortunate that so many workers blindly followed the rules, because Koch himself quickly realized that in certain instances all the conditions could not be met. . . . Thus, in regard to certain diseases, particularly those caused by viruses, the blind adherence to Koch’s postulates may act as a hindrance instead of an aid." Rivers proposed his own set of variants to establish causality when few techniques were available to work with viruses:
- a specific virus must be found associated with a disease with a degree of regularity
- the virus must be shown to occur in the sick individual not as an incidental or accidental finding but as the cause of the disease under investigation. causal relationship between a virus and a disease
This postulate must be read carefully. It does not say that the microorganism is not found anywhere in a healthy animal; it means that the organism is not found in the same type of tissue of a healthy animal. For example, the bacterium Escherichia coli is commonly found in the intestines of a wide range of healthy animals. E. coli in the blood or brain, however, is definitely not compatible with health.
Again, experience has led to modification. With the techniques available to Koch, there were several conditions where the first postulate could not be fulfilled. Koch could not, for example, fulfill the first postulate for viruses, before the invention of the electron microscope. Even more complex is current knowledge that many pathogenic organisms can exist in a subclinical state, producing symptoms only years after infection, such as the human immunodeficiency virus or Mycobacterium leprae.
A variant here is that the host never develops symptoms, but goes into a carrier state, shedding organisms that cause active disease in others. The classic example is Salmonella typhi and "Typhoid Mary".  The organism of most interest to Koch himself, M. tuberculosis, exists in a subclinical state in perhaps most of the people infected with it, although many can infect others.
Yet another variation, where a variation of the postulate can be satisfied by nucleic acids and proteins but not the presence of the organism, is with a disease where the pathology is produced by a toxin excreted by the bacterium. In the case of tetanus, the site of the infection where Clostridium tetani bacteria grew can be far from the central nervous system. Botulism can be even more difficult, because there may be no bacterial growth at all in the host; a lethal dose of toxin came from long-dead bacteria in improperly preserved food.
This originally needed to be modified to allow the definition of "culture" to include living tissue cultures, eggs, etc., in the laboratory. Under this modification, the postulate works for viruses and other pathogens that cannot live and reproduce without the assistance of a host cell.
Ethical issues arise here when an organism, associated with an incurable or lethal disease, is only found in humans. In some cases, such as leprosy (also known as Hansen's Disease, an alternate host was found: the armadillo.
For other diseases, where it was unethical to inject it into humans, Koch's postulates have still been fulfilled. Human immunodeficiency virus has been found in laboratory and healthcare workers, following accidental inoculation from a confirmed "step 3" sample. 
Even though Koch himself did not appear have authored the fourth postulate, it has proven useful and adds further evidence of causality. 
Laboratory infections with HIV demonstrate the fourth postulate, if the genetic material extracted from the now-ill healthcare workers is consistent with the genetic material in the original sample.
- Robert Koch, 1843–1910, vol. Contagion: Historical Views of Diseases and Epidemics, Harvard University Library Open Collections Program
- Fredricks, David N. & David A. Relman (January 1996), "Sequence-Based Identification of Microbial Pathogens: a Reconsideration of Koch’s Postulates", Clinical Microbiology Reviews 9 (1)
- Nova, The Most Dangerous Woman in America, Public Broadcasting Service
- Fredricks & Relman, p. 20
- National Institutes of Health, National Institute for Allergy and Infectious Disease (September 1995), Koch's Postulates Fulfilled: The Relationship Between the Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome
- Fredricks & Relman, p. 19