In this study I set out to describe and refine the method by which we determine the origin of ancient oak timber using dendrochronology. Many aspects needed to be discussed, from the point of view of developing a method to determine provenance to a more local level than previously. The development of the method is initiated on the basis of tree-ring data from living trees, where a set of procedures and rules could be defined. However in the subsequent application of the method of determining the origin of the timber from ancient structures, consideration has to be taken of problems of the historical and/or archaeological context. A series of case studies is presented where the dendrochronological analysis of oak timber from shipwrecks and barrels found in an archaeological context are described, and an attempt is made to determine the origin of the timber. A varying level of detail is applied in each case, depending of the number of samples that are analysed, and depending on the nature of the results that emerge. In some cases a clear provenance to a local level can be attained, while in other instances we can identify the timber origin only to the regional level. The main points that have emerged can be summed up as follows. It has been shown that we can refine the provenance determination technique by testing the tree-ring curve from a structure at three levels. The first level test checks the curve against master chronologies. The second level test is where we test the treering curve with site chronologies. At the third level we check the curve with single tree-ring measurements or indices. In the dendrochronological provenance determination exercise, just as tvalues of greater than 3.5 are interesting when looking for the date of an object, tvalues greater than c. 9.00 are interesting when looking for the provenance of the object. But more importantly, as in dendrochronological methodology, where a date for a tree-ring curve is also checked visually before a position is accepted, the distribution of the correlations and the overlap and replication of the well matching sites is examined before a provenance is suggested. In other words we must look at the geographical distribution of the correlation results in every test. In keeping with a dendrochronological term where an undated chronology is called a ‘floating chronology’, we might refer to the dated but transported chronologies (the panel painting data, the Norwegian timber abroad, site chronologies identified as not native to the area in which they are found) as ‘geographically floating’. We know that timber transport increased over the period dealt with here, but forests, woodland and trees still grew and were utilised locally. It is this combination of usage of local and imported timbers for different uses that allows us to be able to map the movement of timber. As oak timber is easier to work while still ‘green’, preparation at source is necessary, and thus the decision as to what use the timber will be put to might already have been made at source. While the market for the oak panelling that we see in the 14th and 15th centuries might be reliable and regular enough that the production of this timber product could have been a standard activity, it is possible that the preparation of timber for shipbuilding was carried out to fill specific orders, and not as a routine timber product. We can begin to conclude that the transport of bulk oak has to go hand in hand with other lighter timber species, especially when rafting, as the oak timber alone will not float. Oak worked into planks and boards etc make them far more easy to handle, and thereby possible to export on a large scale, while substantial oak timbers, transported over long distances, are a relative rarity. All in all it is logical that if oak is available nearby chances are that it is used, rather than going to all the trouble and expense of using long-distance transport. So the conclusion is that the predominant practice was the use of local oak. Imported oak being the exception, not the rule. It is not until the 16th century that we begin to see the necessity for the transport of oak, and this occurs for those regions which run out of native resources. In this period also we see the increasing dominance of conifer as the main structural timber for building. For the purposes of identifying the occurrence of the transport of timber, as a raw material for shipbuilding, it is shown that the analysis of samples from several timbers of varying form and function in a ship’s structure, bring us nearer the true picture of the timber origin, and the region where the ships were built, which are, by the 15th century, not necessarily one and the same thing! Indeed the pattern emerging seems to point towards the start of the 15th century as the point where, at least in the archaeological record, we see that timber for ship building is shipped to a ship building site some distance from the site where the timber was harvested. In light of the frequent appearance of ancient oak from archaeological sites and from panels and inventory in historic buildings in Western Europe, which shows by dendrochronology to have an eastern Baltic origin, more tree-ring data for oak from the Polish but also from the other Eastern Baltic countries would allow more detailed information of this extensive historic timber trade (Wany 2002; Haneca et al 2005). Clearly, continuing cooperation with dendrochronology laboratories from the underrepresented regions will be an enormous asset for the provenance determination of ancient oak. It can be seen here that when sampling for dendrochronological analysis there is enormous potential for the recording of the types of timber utilised over time, in historical buildings and in the remains of construction found in archaeological excavations. With the possibility of precise felling dates and a review of the quality, dimensions, conversion and tree-age of timbers, we would come towards a detailed picture of the timber in terms of resource availability through time. Not only could we identify instances of imported timber by provenance determination, we could also identify trends in the availability of building timber. This discourse would though have to take into account the different status or social context of the buildings or other constructions, for which the timber is used. Account should be taken for the possibility that the type of timber used in any construction is not necessarily reflecting timber availability, but rather the choice of specific materials with specific qualities.