The production and inspection of food products need reliable chemical analytical methods, to detect whether or not a product meets requirements. Ideas about what constitutes ‘reliable’, and how to measure reliability, have changed drastically during the past few decades. There has been a gradual move away from the ‘I am doing well’ approach, through the use of ‘collaborative studies’ and ‘criteria’, towards ‘quality assurance’ and ‘total quality management’ by means of concepts such as good laboratory practice and accreditation. The probability of obtaining false-positive results and the probability of obtaining false-negative results are useful parameters for measuring the quality of an analytical system. For inspection procedures, in addition to a maximum tolerance, a decision limit and a minimum detectable inadmissible content are also useful. Measuring the ‘quality’ of analytical methods in the case of more complicated situations, such as multivariate or qualitative analyses, still remains difficult.