theoretical...i have pressure vessel (box with one side open) and a
rectangular plate bolted around the open face of the rectangular box. 
I have numerous identical bolts around perimeter of plate sealing box.
 If box is pressurized which bolts will fail first or are they equally
loaded?  How can i calculate tensile load on each bolt, knowing
internal pressure (hand calc)?

Assume your plate bulges in the middle under the pressure.

Then depending on the geometry some bolts on a square plate may be
more stressed than others because of their distance from the centre of
the bulge.

Are they bolts?  or Welded studs?  (do they go through both walls or only the door)

Does the threading pass through the seal?

What steel/material is the vessel/door made of?

Is there a gasket?

Is the door area curved or flat? sphere?

Steam?

Hard to see how identical bolts could be uniformly loaded.

They equallyloaded.
The tensile load on each bolt=[P(internal pressure )*acreage]/bolts number

i guess i'm late to the party, but comments are still being accepted,
so here's my two cents, even if no one ever reads them:

it depends on if you define failure as "first yield" or "ultimate load".

if the plate is rectangular, when everything is elastic and nothing
has yielded, some bolts will be more heavily loaded than others.  in
general, the plate will span in both directions, but the plate will
"want" to span the shorter distance, and "more of the load" (i'm
allowing myself a little fuzziness in my terminology) will go to the
longer side of the rectangle, and the bolts in the middle of the
longer sides will feel more load.  the greater the ratio between the
lengths of the two sides, the more the load will mainly go to the
longer sides.

however, if you can live with a little yielding in plate or bolts, the
additional load from a further pressure increase will redistribute
itself to the other bolts.  how much redistribution you will get
before something (plate or bolt) fails depends on some of the
questions redfoxjumps asks.  it will not get to the perfectly even
distribution that anglefreud mentions, but it might get semi-close. 
the way this is usually analyzed is called "yield line theory".

if you want a hand calc, for either method, a good starting point
would be to get a copy of timoshenko's plates and shells (an
engineering library at a university will likely have a dozen or so
copies).  i'm sure he's solved this case as an example for both the
elastic solution and the yield line solution.

one final warning; if it turns out the bolts yield before the plate,
then they lose stiffness, your boundary conditions change, and i
seriously doubt you'll be able to do a hand calc.  there are lots of
analysis programs out there that could do this, but they cost a LOT
more than $2.

-cab