The decision on conditional release from prison (Art. 86 89 SCC) is one ofthe most important decisions in the Swiss criminal justice system. At stakeare both the freedom of the convicted person, who is to be given a chance ofreintegration by the conditional release, and the security of the general public, which must be protected against the commission of further criminal offences.Despite its significance forconvicted persons and the society as a whole, thepractice of conditional release is scarcely discussed in Swiss literature: Whatis the procedure for conditional release? How do the opinions of the parties involved in the procedure influence its outcome? How do the persons responsible for conditional release exercise the discretion left to them by the law? The clarification of these issues seems all the more important giventhat in recent decades the number of conditional release decisions has decreased and there are considerable cantonal differences.The focus of the work is on the process of decision-making by the responsible authorities in Switzerland, in particular the implementation of theright to be heard, as well as the criteria which influence the decision on theconditional release of offenders. The research is based on a representativesample of 943 criminal files from the cantons of Berne, Fribourg, Lucerneand Vaud, which are evaluated using statistical methods (logisticregression).The results of the investigation show that the responsible authorities selectonly a few aspects from a large number of case characteristics that largelydetermine their decision-making process. The decision is strongly influenced by the residency status of the convicted person, his criminal record and theopinion of the prison authorities. The legal prognosis is therefore primarilybased on static factors from the past, whereas dynamic factors which theconvicted person or the involved actors could influence are more secondary.In addition, it emerges that the temporal and cantonal differences in the release rates do not relate to a differing prison population, as it most often assumed, but to a different way of implementing the law and appreciating theprofile of prisoners following a more restrictive or liberal understandingand practice of the release decision depending on time and canton.The study shows as well that there is no uniform practice for the procedure:the use of risk assessment tools, the opinions of the prison management or the organisation of the right to be heard are extremely disparate and therefore the right of the sentenced person to an equal and fair trial is not guaranteed to the same extent everywhere.The work concludes with a legal classification of the findings as well as withcriminal policy considerations and proposals for a more precise reformulation of Art. 86 SCC. These should contribute to a more harmonious and broader application across cantonal borders and therefore strengthen conditional release as an efficient instrument of crime prevention.