David asked: > I have two versions of the same family tree that I want to merge... Hi David, Others have made good suggestions. I agree that the only reliable way to combine these two trees is manually to review each person and add/move/copy into version A the extra/missing information in version B. I agree with Terry that since you want to have a single family tree, I would first start by creating a separate dataset which has only one tree, and add/modify/move/copy into that dataset the desired information from the other. As I note in a recent reply to a similar question, you cannot "merge" two people from different datasets. You can only merge two people who are in the same dataset. But you can copy. I prefer to do most of the combining manually rather than using copy, as I have very personalized data entry methods and standards. You can have these two versions as two datasets in one project so as to be able to do move/copy. But my trick is to (also) have a second separate "from" project containing the dataset being copied "from". It could contain only the "from" dataset, or could be a copy of the original two-datasets project. You can have open (at most) two separate copies of TMG at the same time. So I have each project open in the two separate windows of the two copies of TMG. I treat the "from" project as "read only" and do all the add/modify/move/copy into the one dataset in the other project. You can even use the Merge feature or Find Duplicates in the "from" project to see their original unchanged states for comparison. If you do choose to merge two projects or datasets, be very careful of any customizations you have done in either, such as custom tag types, custom source templates, etc. > Can someone point to a guide or a discussion ... > for merging the two trees? In addition to Terry's Tips which I earlier recommended reviewing first, my on-line book has an "Import/Export" chapter with a section on handling customizations during merging of two datasets: http://www.mjh-nm.net/IMEXPORT.HTML#MergingDataSets That chapter is a little bit out of date as I have not (yet) added some changes in TMG in this area which were introduced in TMG V8, but it should give you ideas and cautions concerning merging. I include the step-by-step actions I find useful to do first to each dataset while they are still separate, and the sequence of actions I do after the two datasets are merged into a single dataset. There is also a separate section within this chapter on using "merge" for combining people, with my suggestions on the order various actions should take if you choose to "merge" people and families: http://www.mjh-nm.net/IMEXPORT.HTML#MergingPeople This chapter also includes comments about merging locations and merging sources, if those are relevant to you. Finally my separate page of bugs which remain in TMG version 9.05 includes three obscure bugs dealing with merging. They probably do not apply to you, but are mentioned for any others reading this thread: - Problems with merging datasets with Duplicate Tag Type names in different Tag Type groups http://www.mjh-nm.net/BUGS.HTML#B33 - Merge candidates may show non-primary names for parents with name variations http://www.mjh-nm.net/BUGS.HTML#B62 - Merging a dataset containing a Flag which does not exist in the receiving dataset will lose those flag values for the merged people http://www.mjh-nm.net/BUGS.HTML#B92 Hope this gives you helpful ideas, Michael