Jmol is a pretty good free software to use. It's not as powerful as some of the paid packages, but Jmol has a pretty decent library of compounds you can use for referencing,.
Apart from choosing a viewer and measuring the distances you might want to consider the different conformations these molecules can adopt. You could generate several conformers by computational methods and then calculate a mean?
Also remember that pdb files contain the coordinates in plain text. If you write a small script you can just calculate out the atom distances yourself really quick. A bit more comfortable would be to use a library such as biopython or similar which provide a pdb parser and methods for dealing with coordinate files.
1. Use Molden or to generate molecular geometry in cartesian coordinates or internal coordinates (Z-matrix). Molden is free software and is very handy: http://www.cmbi.ru.nl/molden/
2. You may want to generate geometries for several stable conformers.
3. Once the geometries are generated, you may optimize geometry using free softwares like Gamess or HONDO. If you have access to paid software like Q-CHEM, ACES, CFour, NWChem, Turbomol, Gaussian you may choose any one of them as well.
4. For geometry optimization, use at least DZ(d,p) basis set. If you can afford bigger basis, go for TZ(2d,p) or TZ(3d,2p), etc. and the method no less than MP2 or suitable DFT XC functional.
5. Upon successful optimization, the optimized geometries will be printed in the respective output files. Copy these geometries to different files and view them one by one using Molden. You can measure the interatomic distances, angles, dihedral angles, etc very easily using molden.
6. From the geometric parameters of multiple conformers, you may take mean if you want to.