The basic rule for correct design of a sample plant is that all parts of the material being sampled must have an equal probability of being collected and becoming part of the final sample for analysis. However, this rule is often ignored in the design of sample plants to save capital costs. This is false economy, because the samples taken are likely to be seriously biased and the precision may be compromised. The introduction of bias is particularly serious, because no amount of replicate sampling and analysis will reduce bias once it is present and it is often quite difficult and expensive to subsequently rectify the problem due to space limitations around the primary sampling location. This is also true of secondary and tertiary sampling stages, particularly in terms of the relationship between particle size and retained sample mass. The best strategy is to seek expert sampling advice and design the sample plant correctly at the outset.