The Biostator makes it possible to perform glucose clamp experiments almost automatically. Thus, blood glucose concentrations can be maintained at (or close to) a target level with substantially less effort than with the manual clamp technique. The automatisation also avoids a potential bias on the part of the investigator. However, as with the non-automated manual clamp technique, blood glucose concentrations do not remain exactly at the target value, as they show a considerable scatter around the target value. This scatter is generated by the time constants of the Biostator, i.e. the whole closed-loop system, and the autoregressive properties of the glucose clamp algorithm used. In order to describe the quality of glucose clamps over time more precisely, "cumulative sums" as an alternative to the usual coefficient of variation can be used. Practical work with the Biostator is burdened with technical difficulties and considerable costs in comparison to the manual clamp technique. Deficits concerning data storage and presentation capability of the Biostator have been overcome by an appropriate programme for an external computer. The use of the Biostator for the glucose clamp technique is not mandatory, but, the use of this machine makes it possible to perform glucose clamp studies under standardised and reproducible conditions.