Lidar returns from clouds include successive scattering order contributions. Under several assumptions, they can be evaluated by a simple analytical model, where only forward and backward scattering events are considered. Such a model has been developed, which moreover accounts for realistic receiver/emitter characteristics. Two experimental settings, enabling multiple scattering contribution identification, are considered. They are referred as <<2-FOV>> and <<off-axis>> methods. The latter requires no special additional feature to common lidar devices. Both are examined through a simple simulated case, and two measurements performed on cirrus clouds layers. 2-FOV method leads to an accurate estimate of ice crystals diffraction peak width. Off-axis approach, although theoretically efficient and experimentally less requiring, proved to be rather critical, and has to be properly improved.