This work deals with the characterization of two-phase flows based on fractal techniques in order to develop objective flow regime indicators. Fractal analysis has been used previously for studying turbulent flows and it is a natural way of describing highly irregular fluctuations in two-phase flows. The fractal dimension of measured time series have been evaluated by Higuchi’s method. It is shown that the error of the linear fit of the fractal dimension is a sensitive indicator of the changes in the flow regime, while the fractal dimension value itself is less suitable for flow regime identification. The developed method has been applied to the evaluation of two-phase flow experiments at the SIDAS boiling loop at Delft University of Technology, the Netherlands.