Before microdialysis (MD) can be introduced into the clinic as an improved method of cerebral monitoring, certain ethical, methodological and clinical factors must be considered. Access to the brain for probe insertion is offered by craniotomy or by routine intracranial pressure (ICP) monitoring and the additional lesion is minimal. Care must be taken that the two devices do not interfere with each other. In contrast to ICP monitoring, MD provides information about multiple aspects of brain metabolism. We can monitor either still intact tissue to prevent additional damage, or injured brain to decide on and control therapies. The parameters used must reflect pathological changes an early stage, and the analysis should be available on-line or immediately after sample collection. The effects off factors such as tube length and flow rate on the behaviour of the chosen parameters (in our case on-line pH, radical scavengers and uric acid) in the MD set-up must be investigated in vitro and in animal models before use in the clinic. The range of non-pathological values of parameters of interest in human brain should be known For this purpose we took measurements during an extracranial-intracranial bypass operation, and were able to compare values with those in a severely damaged brain. The mutual chronology of parameter changes and clinical events must be clear. Future aspects include the use of low-flow methods offering nearly 100% recovery, improved analytical methods, and combination of MD with other monitoring methods to obtain more exact information.