Classical Particle Image Velocimetry (PIV) uses two representations of the particle image distribution to determine the displacement of the particle image pattern by spatial cross-correlation. The accuracy and the robustness are however limited by the fact that only two representations at t and t +?t are present. Thus, only a first order approximation of the velocity can be estimated. To enhance the precision in estimating the flow velocity, multi-pulse or multi-frame techniques were already investigated in the early days of PIV as summarized by Adrian (1991) and Hain and Kähler (2007). Today with the increasing power of high repetition rate lasers and enhanced sensitivity of the digital cameras it is possible to have a time-resolved sampling of even aerodynamically relevant flows, were the particles are much smaller than in water flows. The easiest sampling scheme is the equidistant temporal sampling of the particle distribution such that a robust displacement estimation between successive frames (1+2, 2+3, 3+4, …) is possible. This so called TR-PIV does not only provide the possibility to follow the evolution of flow structures, but offers the ability to strengthen the data processing by using information from more than two frames (e.g. Hain and Kähler, 2007). Within the AFDAR-project (Advanced Flow Diagnostics for Aeronautical Research funded by the European Union) different approaches to evaluat time-resolved image series were developed by the different groups. The current contribution focuses on the comparison of the algorithms that were developed within the AFDAR project by the partners of the consortium. To verify and validate the performance of the different algorithms a short image sequence of an experiment on the flow over periodic hills (ERCOFTAC test case 81) was provided to all partners and evaluated with the current version of the algorithms.