Dan Lilja says that for a 30-338 cal, roughly 25-30 FPS gain for every inch of barrel. When I run the numbers in QuickLoad for different charges etc., I find variance of 35 FPS/ inch barrel or so down toward the 24"-26" lengths, and up towards the 33"-34" those variances have dropped to about 20-25 FPS/ inch barrel for a given charge weight. Having said that, I have a 34" barrel on my 338LM, and can use RL-50 to get the best velocities for a given OBT, with pressures of 57,000 PSI or so, instead of pushing possible pressure situations close to 60,000 PSI. I like the method of finding an accuracy node, and getting there with OBT calculations, and then using a slow powder in a long barrel to get that----barrel lengths change your OBT, but you can calculate OBT, and input barrel length, and then solve for charge weight to get that OBT---then you see the resulting pressure, and I like the pressures I get with the long barrel.