A measurement of the at-power moderator temperature coefficient (MTC) at the pressurized water reactor Unit 4 of the Ringhals Nuclear Power Plant (Sweden) during fuel cycle 16 is analyzed. The measurement was performed when the boron concentration decreased under 300 ppm in the reactor coolant system, by using the boron dilution method. Detailed calculations were made to estimate all reactivity effects taking place during such a measurement. These effects can only be accounted for through static core calculations that allow calculating contributions to the reactivity change induced by the moderator temperature change. All the calculations were performed with the Studsvik Scandpower SIMULATE-3 code. Analysis of the measurement showed that the contribution of the Doppler effect (in the fuel) was almost negligible, whereas the reactivity effects due to other than the Doppler fuel coefficient and the boron change were surprisingly significant. It was concluded that due to the experimental inaccuracies, the uncertainty associated with the boron dilution method could be much larger than previously expected. The MTC might then be close to -72 pcm/°C, whereas the main goal of the measurement is to verify that the MTC is larger (less negative) than this threshold. The usefulness of the boron dilution method for MTC measurements can therefore be questioned.