Holes quality errors are an undesired but unavoidable consequence in drilling operations. Due to the small dimensions involved in the microdrilling processes, quality measurement and control must be carried out offline, by using microscopy or other high precision measurement devices. This paper presents a study about the correlation between the holes quality and the force signals in the microdilling process of 0.1 mm and 0.5 mm-diameter holes in a sintered tungsten-copper alloy. The surface of the obtained holes was scanned by means of an interferometry microscope and the error of the holes was computed from the scanned data. The three components of the forces were measured during all the drilling process. The behavior of these signals, in three different intervals (tool entrance, forward motion and backward motion) was described by wavelet package analysis. The features having higher correlation with the holes quality error were the average power of the axial component of the forces in the frequency bands of 0~391 Hz and 3906~4297 Hz, during the backward motion. With these features, a statistical regression model was fitted. The main outcomes of this study are the basement for obtaining reliable models for monitoring systems in microdrilling operations.