The feasibility of time-domain NMR curve-fitting methodology for the quantitative determination of TG blend phase compositions was investigated. By studying a range of TG in their different crystal forms, it was shown that the transverse NMR relaxation characteristics of TG differ for the respective crystal polymorphs {α,β,β′}. This enables the TG polymorphism in fat blends to be quantitatively determined by curve fitting. If a liquid phase is present in the blend, curve fitting is able to determine the solid fat content, and the results compare well with those of the accepted NMR methods. The curve-fit method is less hindered by some of the disadvantages of these accepted methods, such as the use of a calibration factor.Paper no. J10095 in JAOCS 79, 383-388 (April 2002).In food technology, the phase composition of fats and oils is an important parameter in process control (1), fat blending properties, and assessments of the quality of various products (2,3). Dilatometry, which is based on the difference in density between the solid and liquid phases, has for a long time been the dominant method to measure the solid fat index (SFI) (4), but this method is laborious and time consuming. Time-domain nuclear magnetic resonance (TD-NMR) has now become the dominant technique, as it offers many advantages regarding speed and ease of operation (2,5,6) and accurate information on the solid-liquid ratio (7,8).The current TD-NMR methods for SFC determination of fat blends have been successfully applied in fat technology for several decades. On one hand, the indirect method (2,7) is considered accurate but not precise, whereas, on the other hand, the "direct" method (2,5) is precise but not accurate and needs a calibration factor (derived from plastic-in-oil samples). In the hands of a skilled operator, the solid-echo (8) method is considered to be both accurate and precise but has never caught on in routine laboratories. Despite its obvious limitation, the direct method has become the most widespread method for SFC determination due to its ease of use and its high precision. However, since the introduction of this direct SFC method in the 1970s, the specifications of commercial TD-NMR equipment have improved dramatically. Also, there is a belief that the current methods do not fully exploit the potential of modern technology.In the detergent area, several applications have been described that show how the phase behavior of surfactants can be assessed by curve fitting of the time-domain data (9). So far, only a few examples are known where the phase behavior of lipids has been assessed (10,11). In recent reports (12,13), it was demonstrated that curve fitting of rapidly sampled transverse NMR relaxation decay curves enabled the quantitative determination of the mesomorphic state of polysaccharides. As the curve-fitting method is not fraught with some of the disadvantages of the current SFC NMR methods, we have embarked on a study to investigate the feasibility of curve fitting for quantification of MG and TG phase compositio...