A quality control procedure is described for biodiversity data compilations based on concepts of Statistical Processing Control (SPC). The technique, called process control sampling, involves continuous inspection of biological materials coupled with ongoing training. The describe d procedure contrasts with others where quality control is confined to acceptance sampling (another form of SPC) and one-off training. The basic laboratory sorting procedure we have developed is described, as are methods for monitoring error levels. These include control charts and taxonomic breakdown charts. In biodiversity data compilations three basic types of errors are identified: misidentification, mislabelling and accidential inclusions. The process control protocol prevents proliferation of controllable errors such as errors of misidentification. It also addresses the problem of prior knowledge of which tax a are likely to be encountered , and which are likely to present problems for particular technicians. Further, uncontrollable errors such as mislabelling and accidental inclusions require continual monitoring in exercises which are large scale and repetitive in nature. A process control methodology has the advantage of ongoing monitoring of technical staff throughout the project, coupled with the ability to tailor feedback training to suit an individual’s requirements. The significance of error levels at higher levels of taxonomic sorting is emphasized.