A key problem that arises when unstructured text is being queried is that of properly recognizing and exploiting geographical terms and entities. Here we describe a mechanism for probabilistic toponym resolution, and our experiments with the new method in the setting of the 2005 GeoCLEF queries and judgments. The new method gives improved retrieval effectiveness on a subset of the topics. Categories and Subject Descriptors H.3 [Information Storage and Retrieval]: Content Analysis and Indexing – Indexing Methods Keywords Geo-spatial information retrieval, toponym resolution