I have a table of over 1 million test score records that basically have a unique score_ID, a subject_ID and a score given by a person. The score range for most subjects is 0-3, but some have a range of 0-4. There are about 25 possible subjects.

I need to produce a score distribution report which looks like:

subject_ID     0    1    2    3    4
----------    ---  ---  ---  ---  ---
1             967  576  856  234  
2             576  947  847  987  324
.
.

So it groups the data by subject_ID, then shows how many times a specific score value was given within that subject.

Any SQL pointers to generate this would be greatly appreciated.