This paperpresents an application of acondition-monitoring system (CMS) based on a polynomial regression model (PRM)to study the influence of heat loss on a wind generator’s temperatures.Monitoring the wind generator temperatures is a significant for efficientoperation, and plays a key role in an effective CMS. Many techniques, includingprediction models can be utilized to reliably forecast a wind generator’stemperature during operation and avoid the occurrence of a failure. PRMs arewidely used in situations when therelationship between the response and the independent variables are curve-linear.Thesetechniques can be used to construct a normal behavior model of an electricalgenerator’s temperatures based on recorded data. Many independent variables affect agenerator’s temperature; however, the degree of influence of each independent variable on the response is dissimilar. In manysituations, adding a new independent variableto the model may cause unsatisfactory results; therefore,the selection of the variables should be veryaccurate. A generator’s heatloss can be considered a significant independent variable that greatlyinfluences the wind generator with respect to the other variables. A generator’s heat loss can be estimated in intervals by analyzing theexchange in the heat between the hot and cold fluid throughthe heat exchangers of wind generators. Acase study built on data collected from actual measurements demonstrates theadequacy of the proposed model.