This paper presents the results of an experimental evaluation of the performance of a dynamic distributed federated database (DDFD). Databases of varying sizes up to 1025 nodes, have been evaluated in a test environment consisting of a 22 IBM Blade Servers plus 3 Logical Partitions of an IBM System P Server. The results confirm that by 'growing' databases, using biologically inspired principles of preferential attachment, that the resulting query 'execute time' is a function of the number of nodes (N) and the network latency (T L) between nodes and scales as O(T L logN). Distributing data across all the nodes of the database and performing queries from any single node also confirms that, where network bandwidth is not a constraint, the data 'fetch time' is linear with number of records returned.