The flow pattern generated in pipes for two-phase gas-liquid horizontal flows can significantly affect the efficiency of chemical process equipment. Since the early 1900's, scientists and engineers have investigated various methods for characterizing the flow patterns generated for various two-phase chemical systems. A notable effort was made by Baker to quantify physical fluid properties and correlate them to observed flow patterns in pipe flow. The primary objective of this research was to develop a flow pattern recognition system, for industrial implementation, that is suitable for characterizing transient two-phase flow patterns. Since the advent of computer and system control equipment, researchers have focused on quantifying flow characteristics by analyzing pressure fluctuation because the natural tendency of multiphase flow is to pulse as the fluid phases mix while moving along the pipe length. Recent developments in digital signal processing provide new methods for characterizing transient signals. A novel approach using wavelet transform analysis to study pressure fluctuations in two-phase flow systems has been developed in this research. Experimental development includes the design of a baseline system and a flow pattern recognition system. (Abstract shortened by UMI.).