This paper illustrates that sampling problems and difficulties are as common in an R&D environment as they are in an operating plant. Researchers need to be aware of the potential influence of poor sampling practices and when feasible determine the sampling and analytical precision at the start of any experimentation. Specific problems that have been encountered are discussed and the procedures implemented to overcome these difficulties are described. Examples from projects include the cyanidation of a Au/Ag ore, column leaching of refractory gold and oxide ores, the bioremediation of contaminated sites and the flotation of ore containing native copper.