My accuracy varies greatly with my AR-50 when using API's and M-33's.  Some days I can put three shots in a quarter @ 100yds, other days that group extends out to 3 inches with no change in load.  I do weigh and group the bullets and use only the same year LC brass per batch but yet accuracy is wild.  The current load I'm using is 212 grains of virgin WC867 behind API's and M-33's with  velocity being in the low 2,700's ten feet from the brake.  I'll wager that someone here can save me some time and tell me if it's the bullets or the powder.  I'm thinking the powder is temperature sensitive and causing the problem.  It this the case?

I have never reloaded much but have a great memory for articles I have read.

Things to consider:
Are the velocity readings the same on those different range days you describe?
If different were they within the same spread range amount?


Were the bullets loaded on the day of shooting some times and not the other? Supposedly used brass can relax over time after the crimp causing pressure changes which is why some long range bench shooters reload at the range when they are shooting.  Annealing the cases may help with that issue (good youtube clip showing home made annealer for 5.56) which leads to the next question.
Do you anneal your brass?

Just some things to consider.  Good luck.

Absolutely....I run them through the sizer die as many times as necessary.

872 powder is known to be temperature sensitive, and different lots will fire with different pressures. Watch for clumps in your surplus powders.

I did a test years ago heating rounds loaded with 872 and SA Ball - heated the whole batch to 125 degress F and it helped to minimize velocity variations.

V.V. 20N29 powder is also very temperature sensitive, which is why we heat loaded rounds to 130 degrees for best accuracy. The cost of 20N29 is prohibitive these days which is why most accuracy reloaders are using Reloder50 or Hodgdon H50BMG.