The present invention provides a technique that enables an image communication apparatus capable of performing voice communication and T.30 in-band FAX communication over the IP network to positively detect a CNG signal and achieve switching to FAX reception in the answerphone connection mode and the FAX/TEL-switching mode. The communication apparatus includes a MODEM that controls facsimile communication and a SLIC that controls a telephone. A selector includes switches for connecting the IP network and the MODEM, and switches for connecting the IP network and the SLIC. A CPU controls the switches of the selector such that signals from the IP network are supplied to both of the MODEM and the SLIC.