The infrastructure involved in providing medical services is complex and expensive, and encompasses both human resources and equipments; therefore, it needs an adequate resource management to attain profitable results. As stated by Spyropoulos (2000), a hospital infrastructure is composed of: (a) Human resources: physicians, nurses, administrative personal, technicians for equipment maintenance, etc. (b) Intensive therapy units, with an expensive infrastructure. (c) Surgery room, with dedicated equipments for several procedures. (d) Specialized laboratories: X-rays, ultrasound, tomography, magnetic resonance, etc. (e) Auxiliary infrastructure: ambulance for emergency transfer, patient’s rooms, pharmacy, restaurant, etc. In recent decades many tools with the aim of providing efficient management of this infrastructure have been proposed. Oddi and Cesta (2000) considered that managers of medicohospital facilities are facing two general problems when allocating resources to activities: (1) to find an agreement between several and contrasting requirements; (2) to manage dynamic and uncertain situations when constraints suddenly change over time due to medical needs. This paper describes the results of a research aimed at applying constraint-based scheduling techniques to the management of medical resources. A mixed-initiative problemsolving approach is adopted in which a user and a decision support system interact to incrementally achieve a satisfactory solution to the problem of allocating resources to medical activities. The authors claim two main contributions of the paper. The first one concerns the domain modeling. The medical problem is represented as a Constraint Satisfaction Problem (CSP) (Tsang, 1996), hence described as a set of variables and a set of constraints on the values of the variables. A solution to the problem is a variable assignment which is compatible with all the constraints. Two main objects are represented in this schema: medical protocols and resources. The constraints are classified as relaxable or nonrelaxable. The solution represents an agreement between different and contrasting goals by reducing the total amount of violations of non-relaxable constraints. A second contribution is the introduction of a new solution algorithm, in which two types of algorithm are integrated: a greedy procedure to create an initial solution and a local search method to improve the initial solution with respect to the amount of violations contained in it. The local method used is tabu-search. Valouxis and Housos (2003) presented a detailed model and an efficient solution methodology for the monthly work shift and rest assignment of hospital nursing personnel. A model that satisfies the rules of a typical hospital environment based both on published research data and on local hospital requirements is designed. A hybrid methodology that utilizes the strengths of operational research and artificial intelligence was used for the solution of the problem. In particular, an approximate integer linear programming (ILP) model is firstly solved and its solution is further improved using local search techniques, as tabu-search strategy.