In variographic analysis the nugget effect provides an estimate of the total process measurement system error. The value of variographic analysis can be further increased by differentiating the analytical error from the sampling errors. The ratio of the minimum practical error (MPE) to the analytical error provides an estimate of the suitability and performance of a sampling system since the analytical error would still remain even if sampling errors could be completely eliminated. The sampling and analytical errors for four systems used to obtain 1-D lots of pharmaceutical powder blends is presented. The first studies were conducted with blends moving over a conveyor belt and with the feed-frame of a tablet press. The researchers then developed and patented a new stream sampler and chute for sampling and analysis of pharmaceutical powder blends. Near Infrared or Raman spectra were obtained as the powder blends flowed or moved and used to determine the drug concentration in the blends. Even though these methods do not require sample extraction and sample preparation in a laboratory, they are still subject to sampling errors which were estimated through the variographic analysis. The MPE was compared to the analytical error for these four systems. The results obtained show that it is often possible to reduce sampling errors to less than ten times the analytical error. These studies represent the first efforts to estimate and reduce sampling errors in the analysis of the powder blends used to manufacture the tablets that many patients take daily.