A neurofuzzy methodology for online nodal load prediction is introduced that exploits the power of artificial neural networks (ANN) and fuzzy logic. ANNs are used to capture the power consumption patterns specific to a customer, while a fuzzy logic module detects departures from equilibrium (that is, previously established consumption patterns). The fuzzy-logic-based (FL) module (called PROTREN) performs signal trend identification. The proposed methodology improves the adaptability of the forecasting system to sudden changes or special events that may influence the load by temporarily distorting the general pattern and thus rendering the load signal highly unpredictable. Experiments have been performed to verify the effectiveness of the new methodology. Results show that the methodology has a better performance than those using traditional forecasting methodologies, especially when special events influencing the load occur.