This International Standard specifies a method of determining whether an inferential system for natural gas analysis is fit for purpose. It can be used to either 1. to determine a range of gas compositions to which the method can be applied, while satisfying previously defined criteria for the maximum errors on properties, 2. to evaluate the range of errors on a property when analysing gases within a defined range of composition. It is assumed that a) the analytical requirement has been clearly and unambiguously defined, in terms of the range of composition to be measured, and, where appropriate, the range of properties which may be calculated from these measurements b) the analytical system is intended to be applied to gases having compositions which vary over ranges normally found in gas transmission and distribution systems. If the performance evaluation shows the system to be unsatisfactory in terms of the error on the inferred properties. The instrument shall be reviewed to assess where improvements can be obtained. This International Standard is applicable to inferential systems which do not measure all of the individual component amount fractions, but instead infer the physical properties by the direct measurement of other properties. Certain components may be measured directly to help improve the accuracy of the inferred properties. Performance evaluation of an analytical system should be performed following initial installation to ensure that errors associated with assumed response functions are fit for purpose. Thereafter, periodic performance evaluation is recommended, or whenever any critical component of the analytical system is adjusted or replaced. The appropriate interval between periodic performance evaluations will depend upon both how instrument responses vary with time and also how large an error may be tolerated. This first consideration is dependent upon instrument/operation; the second is dependent on the application. It is not appropriate, therefore, for this standard to offer specific recommendations on intervals between performance evaluations