Abstract: Computational grids provide mechanisms for sharing and accessing large and heterogeneous collections of remote resources such as computers, online instruments, storage space, data and applications. Resources are identified based on attributes. Resource attributes have various degrees of dynamism, form mostly static attributes, like operating system version, to highly dynamic ones, like network bandwidth or CPU load. In this paper we propose a P2P architecture for resource discovery in a large and dynamic collection of resources. We evaluate a set of request-forwarding algorithms in a fully distributed architecture, designed to accommodate heterogeneity (in both sharing policies and resource types) and dynamism. For this, we build a test bed that models two usage characteristics:(1) resource distribution on peer, that varies in the number and the frequency of shared resources; and (2) various requests patterns for resources.