To present an example of how to study and analyse the clinical practice and the quality of medical decision-making under daily routine working conditions in a radiotherapy department, with the aims of detecting deficiencies and improving the quality of patient care. Bimonthly audits (6 with a duration of 4-6 hours in each institution) were carried out by 3 auditors from the evaluating departments; they reviewed a total of 452 cases in Department A, and 265 cases in Department B. During the ongoing external audits, the qualifying parameters were (1) the sound foundation of the indication of radiotherapy, (2) the conformity to the institution protocol (3), the adequacy of the choice of radiation equipment, (4) the appropriateness of the treatment plan, and the correspondence of the latter with (5) the simulation and (6) verification films. Various degrees of deviation from the treatment principles were defined and scored on the basis of the concept of Horiot et al. (10), with some modifications. The action was regarded as adequate (score 1) in the event of no deviation or only a small deviation with presumably no alteration of the desired end-result of the treatment. A deviation adversely influencing the result of the therapy was considered a major deviation (score 3). Cases involving a minor deviation (score 2) were those only slightly affecting the therapeutic end-results, with effects between those of cases with scores 1 and 3. Non-performance of the necessary radiotherapeutic procedures was penalized by the highest score of 4. Statistical evaluation was performed with the BMDP software package, using variance analysis. Despite the comparable staffing and instrumental conditions, a markedly higher number (1.5 times) of new cases were treated in Department A, but with a lower quality of radiotherapy, as adequate values of qualifying parameters (1)-(6) were more frequent for the cases treated in Department B (85, 94%, 83%, 28%, 42% and 81%) than for those in Department A (67%, 83%, 88%, 26%, 33% and 18%). The responsible division (including staff and instrumentation), the responsible physician and the type of the disease each exerted a highly significant effect on the quality level of the treatment. Statistical analysis revealed a positive influence of the curative (relative to the palliative/symptomatic) intention of the treatment on the level of quality, but the effect of the first radiotherapy (relative to the second or further one) was statistically significant in only one department. The external audit related to the provision of radiotherapeutic care proved feasible with real valuation of the staff's activity.