Replacement stands for any method, approach and strategy not using living animals. The most informative replacement methods are those predicting best human pathophysiological effects. The methods are applicable to basic and applied biomedical science, education and toxicology covering models predicting effects in healthy and diseased tissues and organs.
The most important criteria for a non-animal test method are that it predicts the human effect (human mechanism or end point) sufficiently and is reliable. The most useful methods are those that have been characterized and standardized/validated, but scientifically justified methods and models might be very useful, too.
University researcher or research groups develop models and methods that are mostly applied for their own use but could be widely used by others or their expertise could be broadly leveraged. Universities may also have research services available. The contract research laboratories and industry provide services with standardized or tailor-made tests for clients´ needs.