The authors demonstrate that in order to increase the oil production at a late stage of operation of oilfields, a timely identification of the technical state of the sucker rod pumping unit (SRPU) is required. They point out that emergence of a fault in SRPU is accompanied by a random additive noise, which is superimposed on the signal, coming from the force transducer. It was established experimentally that by calculation of the characteristics of the noise, inextricable from a noisy signal, it was possible to identify the initial period of origin and determine the degrees of fault development in SRPU. It was discovered that the noise probability density function could be used as an indicator for determination of the degree of such SRPU faults as plunger sticking, discharge valve leakage (PVL) and discharge pipes leakage, suction valve leakage (SVL), pump pipes leakage, and slackening leading to sucker rod breakage. The authors developed algorithms for calculation of the noise probability density function, its maximum and the coordinates of the inflection points, and proposed algorithms for identifying the degree of SRPU fault by means of the noise probability density function, and the relevant analysis was carried out. A bank was established, comprising a bank of the graphic images and a databank of the discrete values of the noise probability density function corresponding to various degrees of faults.