The use of an internal standard in NMR spectroscopy to determine the strength (assay) of a material incurs several errors. The sources and magnitude of these errors were examined so that they may be minimised. Minimisation of these errors will result in strength determinations of these materials with 95% confidence limits of ±1%.