This research is focused on a comparison of the classic and the strain Experimental Modal Analysis (EMA). The modal parameters (the natural frequencies, the Displacement Mode Shapes (DMSs) and the damping) of real structures are usually identified with the classic EMA, where the responses are measured with motion sensors (mph{e.g.}, accelerometers). The strain EMA is a special approach in the field of EMA, where the responses are measured with strain sensors. Classic EMA is the preferred method, but strain EMA offers advantages that are important for particular applications: for example, the direct identification of Strain Mode Shapes (SMSs), which is important in the vibration-fatigue and damage-identification models. The next advantage is that the strain EMA can sometimes be used, for experimental/geometrical reasons, where the classic EMA cannot.There are also drawbacks: mph{e.g.}, with strain EMA only, the mass-normalisation of the DMSs and SMSs cannot be performed. This study researches the theoretical similarities and differences of both EMA approaches.Furthermore, the accuracy of both approaches for the case of a free-free supported beam and a free-free supported plate is investigated. The classic and the strain EMA were performed with a piezoelectric accelerometer and the piezoelectric strain gauges, respectively. The results show that the accuracy of the strain EMA results (the natural frequencies, DMSs and the damping) is comparable to the accuracy of the classic EMA.