Modern power systems are currently operating under heavily loaded conditions due to various economic, environmental, and regulatory changes. Consequently, maintaining voltage stability has become a growing concern for electric power utilities. With the increased loading and exploitation of the power transmission system, the problem of voltage stability and voltage collapse attracts more and more attention. A voltage collapse can take place in systems and subsystems, and can appear quite abruptly. There are different methods used to study the voltage collapse phenomenon, such as the Jacobian method, the voltage instability proximity index (VIPI) and the voltage collapse proximity indicator method. This paper is concerned with the problem of voltage stability, and investigates a proposed voltage collapse proximity indicator applicable to the load points of a power system. Voltage instability is early predicted using artificial neural networks on the basis of a voltage collapse proximity indicator. Different system loading strategies are studied and evaluated. Test results on a sample and large power system demonstrate the merits of the proposed approach. The objective of this paper is to present the application of ANN in estimating the voltage collapse proximity indicator of a power system.