The scientific method is based on the interaction of evidence and theory: the theory explains the evidence, and experiments produce evidence to test the predictions of the theory. Without a theory the significance of evidence can be missed; without experimenting to gain evidence, a theory cannot be tested. Homeopathy is accused of being unscientific, but its whole development has been based on the use of the scientific method, and its practice involves a continual testing of theory against evidence. The actual reactions of a patient to each remedy are assessed in the light of the theory, and if the reaction is unexpected then the grounds for selecting the remedy are reassessed to determine why the expectation was wrong.
A founding principle of homeopathy is that there must be a consistent relationship between medicines and the illnesses they cure. This is summed up as:
A. Substances must be tested in order to discover their unique action on human beings.
B. All the individual’s symptoms (as far as possible) must be established in order to know what must be treated.
C. There must be a clear method of relating the information about the action of substances (A) to the information about what needs to be treated (B) in order to cure. 
In fact there are only three options when it comes to relating these two bodies of information to each other:
1. The first option is to suppose that there is no consistent relationship between the effect of a substance on human beings and the symptoms experienced by human beings. If this is the case there can never be a science of medicine.
2. The second option is to suppose that for a substance to be effective in treating symptoms it must, as nearly as possible, be able to produce symptoms which are the opposite of those it is to treat. The problem is that for many conditions (such as spots or a sore throat, for example) there is no opposite, since they are either present or absent. In addition clinical experience indicates that this approach leads to only brief benefit and a need to repeat the medicine. Historically the case of opium as a painkiller illustrates the point, since the the symptoms returned ever more quickly and more strongly later. A modern example of this is seen in the rebound effect of some drugs.
3. The third option is lto suppose that for a substance to be effective in treating symptoms it must, as nearly as possible, be able to produce the same symptoms as it is to treat. As it happens, clinical evidence indicates that this is the case.: a prime example is the use of quinine for centuries to treat malaria, since the symptoms of quinine poisoning are extremely similar to those of malaria.  This was the example, in fact, which led to the discovery of the principles of homeopathy. Other examples include silver nitrate for neonatal conjunctivitis, coal tar for psoriasis, or gold for arthritis. Furthermore this clinical evidence is entirely consistent with the theoretical requirements of homeostasis.
Observation led Samuel Hahnemann to think that "like cures like", and both experiment and theoretical analysis confirmed his observations, used in accordance with the scientific method. Furthermore the development of homeopathy on this sound foundation has enabled the it to make other significant advances, such as formulating a 'law of cure', being able to identify and provide detailed explanations of different reactions to treatment, and establishing a relationship between illnesses arising from infectious micro-organisms and non-infectious chronic illnesses.
Far from having anything to fear from the scientific mehod, homeopathy is entirely dependent on it for the discovery of its principles, its development and its success.
1. Samuel Hahnemann (trans. William Boericke), Organon of Medicine, 6th edn, manuscript completed 1841, 1st English edn 1921 (Calcutta: Roy Publishing House, repr. edn 1972), § 3 p. 90.
2. ‘Qualaquin (quinine) – Cinchonism’, Doublecheckmd website at http://www.bmj.com/cgi/content/full/312/7023/71, accessed 6 December 2008.