Commercial potato chips offer diverse compositions, while the
nutritional label verification of chips batches by official methods is
difficult. Thus, alternative technology is required. Nutritional label
information is commonly provided as percentages of several nutritional
categories. These correspond to data mathematically so-called
compositional, for which specific statistical methods have been
developed. Near-infrared spectroscopy is a rapid and non-destructive
technique proven as useful in many different applications in the food
industry. However, the need for the specific treatment of the
compositional references used with NIRS methods has been considered, so
far, in a few recent studies only. The present study proposes the
methodology for measuring the nutritional label components of potato
chips fat. Compositional calibration models based on visible and
near-infrared spectroscopy were fitted by partial least squares. They
were used to estimate the composition of saturated, monounsaturated, and
polyunsaturated types of fat, as well as their total fat. This study
proposes a useful compositional method for measuring the components of
the potato chip's nutritional label or its verification.