Renderings

No official orthographic drawings of the Excelsior, or any version thereof, has ever been released. All drawings I have seen have been made from the photos of the model. All model kits and toys have similarly been created from photos. All have a number of errors.

The first drawing comes from the patent filing for a "toy spaceship" (patent D288945, March 24, 1987).

The second drawing was published by FASA as part of their role playing game sourcebook for Star Trek III. They also published the first dimensions for the ship at 467 meters long, 186 meters wide, and 78 meters tall.

FASA also released a 1/3937 scale miniature. One of the few to get most details right, but very flawed in execution. The top of the saucer was flattened and the grove on the bottom was missing - replaced with an indent reminiscent of the Constitution Class.

AMT released a 1/1000 scale model kit in conjunction with Star Trek VI. It featured the correct changes from the first version, but many incorrect details.

Fans first made a near clone in the USS Ingram - NCC-2001 and then Jackill's put out their version, which takes many of its details from the AMT kit.

In both Star Trek VI and Star Trek Generations, a display is visible that provides the first real drawings of the ship, but they are, again, very flawed. (The higher quality Star Trek VI one is a recreation that actually corrects many of the flaws of the original, which shows the original ship configuration, not the Star Trek VI configuration)

The Star Trek Encyclopedia had this image, but again, very flawed. It is obvious this was taken from the original profile view photo without correcting for the perspective distortions. This is probably the source of the incorrect RCS placement and the off-center pylons, something it shared with the AMT kit.

This fan made display is probably the most correct drawing available. It has the phasers and RCS aligned to the grid and the correct grid spacing. It has the generally correct proportions and the pylons are centered. It is not perfect, but it is far more accurate than anything shown above.

These renders by S. Matthew Coles from 2005 and Tobias Richter from 2012 are both a step closer to the studio model. Coles' model is very close to the original configuration while Richter's is very close to the second model by Greg Jein.