Radiosurgery represents a minimally invasive radiologic method for the treatment of intracranial tumours and arteriovenous malformations. In 1994 the radiosurgical device 'Leksell Gamma Knife' (LGK) was installed in a dedicated logistic environment for outpatient treatments. High quality requirements have to be met for radiosurgery. The target point accuracy taking into account the whole system was shown to be reliably below 0.5 mm whereas the spatial therapeutic resolution was 0.035 cm3. Quality parameters of the dose plan were evaluated for the first 500 consecutive treatments. These values and examples of dose plans were used to emphasize the advantages of the treatment principle with multiple isocenters. An analysis of data in the literature revealed that there is no uniform standard of treatment available in radiosurgery. A highly significant correlation between a risk prediction model for the stereotactic linear accelerator on the one hand and a different model for the LGK on the other could be shown. This result could be helpful in order to proceed towards a more uniform treatment standard in radiosurgery and to improve overall treatment results.