Hi Jayant, The Chimera fit optimization tool Tools / Volume Data / Fit Models in Maps rigidly rotates and moves the model to increase the average density value at the positions of selected atoms. That is not necessarily what you want. Depending on your map it may push your PDB model towards some high density region even though another position better matches the shape of the map contour surface to the shape of the molecule. We don't have any fitting tool to optimize this shape matching although maybe we should make one. I could imagine taking a molecular surface for the PDB model and trying to position it to best match the contour surface of the EM map. The one choice you can make is which atoms you select for doing the fit. If some subdomain of your model fits the map well you could just select that domain. The fit will move the whole PDB model to optimize just the density at the selected atom positions. If that does not work, the current fitting tool may not be helpful, and a hand placement may be the best that can be done. It is possible to extract just a portion of the map around your hand placed model. http://www.cgl.ucsf.edu/chimera/tutorials/volumetour/volumetour.html#zone Then you could use the fit optimization tool just on that saved piece of the map. But that very likely is not helpful. Your PDB model will just move towards the undesirable position until it hits a boundary where you cut the map off. So I think a hand-fit or using other software (e.g. EMFit or Situs) is the way to go. If you hand place your PDB and then the fit optimization messes it up you can bring the PDB exactly back to the hand placed position with Tools / Movement / Undo Move In general this undoes your previous model movements (remembering a history of I think 100 previous positions). This is so useful that I put it on the Chimera toolbar using Favorites / Preferences, category Tools, checking the "On toolbar" button for "Undo Move" and the Save button on the preferences dialog (so it will be on the toolbar in future sessions). Tom