The dynamics of non-esterified fatty acid (NEFA) metabolism in lactating dairy cows requires quantification if we are to understand how dietary treatments and disease influence changes in body condition (adipose reserves) and the production of milk fat. Recently, Thomaseth & Pavan (2003) presented a compartmental model (Thomaseth model), which employs the pattern of plasma insulin concentrations in humans to predict the dynamic changes that occur in the plasma concentrations of glucose and NEFA during an intravenous glucose tolerance test (IVGTT). The Thomaseth model, or at least a similar model, could have potential application to the field of energy metabolism in ruminants because it would enable the estimation of parameters that describe the rates of whole body disposition of glucose, and the production and utilization of NEFA. In this study we investigated the suitability of the Thomaseth model to describe NEFA and glucose kinetics in 10 lactating Holstein-Friesian cows given a standard IVGTT in early lactation. The Thomaseth model described the general pattern of the NEFA response and, in particular, described the downward-slope and nadir in NEFA concentrations reasonably well. However, it failed to describe the initial latency period (the period before NEFA concentrations decline precipitously), and it could not describe terminal 'rebound' plateau in NEFA concentration. Because of these inherent problems, the parameters of the Thomaseth model cannot be considered to provide accurate estimates of rates of NEFA production or utilization. It is concluded that the Thomaseth model is not suitable for describing NEFA kinetics in lactating dairy cows.