In popular discussions of science, terms such as theory, hypothesis, and even fact are often used quite loosely; in terms of scientific method and philosophy, however, these words actually carry their own specific meanings. In this case, the fundamental difference between a “theory,” on the one hand, and a “hypothesis,” on the other, is that a hypothesis is simply an educated hunch or prediction about the world, which must be tested through experimentation or observation, whereas a theory is a body of such predictions which have consistently held true in the past and can therefore be used to explain current data and to make further predictions.
All theories have their origins in hypotheses. Physical and social scientists refer to their initial argument or suggestion about the way the world works as a hypothesis: in other words, it is an answer to a question, or solution to a problem, which seems to be plausible but hasn’t yet been reliably confirmed through analysis of evidence. The more often and more conclusively a hypothesis is confirmed through scientific tests, the more commonly accepted it will be by scientists working in the relevant field.
At the moment, the “theory” of intelligent design also qualifies as a hypothesis – although it is, at best, debatable whether there are naturalistic means of testing this hypothesis. Other hypotheses are also important in certain areas of scientific research; for example, scientists have yet to confirm the hypothesized existence of dark energy in the universe, but if it does exist (as has been suggested), it would explain a lot of puzzling inconsistencies which currently acknowledged theories are unable to explain, notably the fact that the universe seems to be considerably lighter than our best models otherwise suggest it should be.
Precisely when a hypothesis or set of hypotheses has been accepted long enough to qualify as a theory is a fuzzy matter, at best (an irony given that science in general prizes certainty wherever it is plausible and possible to do so). In general, a theory is commonly accepted as a valid explanation of some facet of the natural world by many or most researchers in a given field, and is accepted as at least a reasonably merited proposition even by those who are reluctant to accept its validity.
Natural evolution, for example, is a currently accepted biological theory. It has been repeatedly tested through observations of the fossilized and other remains of past organisms, and has largely held true. In the process, there have been various amendments to the theory of evolution itself, the most important of which probably involve the role of genetics and DNA (something Charles Darwin, for example, was entirely unaware of). For this reason, the present state of a theory may or may not closely resemble the hypothesis which it originally took the form of.
Even the strongest theory, it should be noted, is still only a theory – that is, it is well-regarded only because it has reliably explained the results of past experiments and observations. New evidence, or new tests (or both), may well overturn even the best-established of scientific theories. For example, Newtonian theories of physics (classical mechanics and the theory of gravity) were considered quite valid until they were overturned by Albert Einstein’s theory of relativity, and are still considered useful enough when applied to relatively heavy, slow-travelling objects.