Estimating the source of a printing process is of primary importance for investigative and forensic purposes. This paper deals with development and application of a method for identifying and broadly classifying the print process from a given print sample, especially official forms and certificates. The developed method was not limited only to electrophotographic (EP) method, but was extended to conventional printing methods like offset and gravure. Various methods were previously tested on EP printed samples, but its use on conventional printing methods are being presented for the first time in this paper. A number of features were extracted from the print samples collected, which includes the various halftone structural transformations that inherently becomes a part of the printing process. Also, the halftone dots that forms a part of the image in any printing process were analyzed using Hough transform and the features thus collected were used to form a reference database. The features in this database formed as inputs to a neural network classifier, whereby the classifier was trained to obtain outputs and classify them into EP, offset or gravure classes. Once the trained network was formed newer test print samples were used to extract the same features and fed into the network for classification. The results obtained clearly showed that the outputs were correctly recognized and hence the method presented is quite promising.