Titanium alloys, largely used for aeronautical applications, are difficult to machine. High cutting forces, chip serration and important tool wear reflect this poor machinability, limiting productivity. One way of improving the machinability of titanium alloys consists of controlling their microstructure. In the present work, the impact of the microstructure of the Ti5553 alloy on chip formation and cutting forces is investigated. For this purpose, a novel experimental approach is proposed. Orthogonal cutting tests are performed on eight different microstructures, which allows studying the impact of the α-phase fraction as well as the size and shape of α particles. Also, an original post processing method based on machine learning provides chip morphological information from images recorded with two high speed cameras. Such information is completed with the cutting forces measured with a dynamometer. In contrast with commonly used approaches, the proposed method is not limited to the formation of a few segments, but uses the full dataset acquired during a test. The results obtained for the different microstructures indicate that no direct link can be established between the cutting forces and their hardness as minimal cutting forces are obtained for microstructures with an intermediate hardness. For microstructures providing low hardness, high cutting forces result from a significantly thick chip. In opposition, for the microstructures leading to high hardness, an important flow stress generates high cutting forces. This study also suggests that chip morphology is primarily affected by the α-phase fraction while the size and morphology of α-phase particles have little influence.