This paper summarizes a set of surveys on ten main French public transport networks including LRT lines: Le Mans, Marseille, Montpellier, Orléans, Reims, Valenciennes, Toulouse, Nancy, Caen and Clermont-Ferrand. This set associates single-line and more complex networks, classical tramways, LRT on tyres and « hybrid » systems (partially guided). The performance of the network is based on (i) the profile of the lines (ramps, curves, etc.), (ii) the speed limits, (iii) and the degree of independence of LRT tracks vis à vis general road traffic (which can generate risk of conflicts and delays). Current travel times (and average commercial speed) are confronted to those announced in the projects. In most of the cases studied, average speeds are lower than expected (generally between 15 and 18 kph). We have tried to determine the reasons of the lower level of performance. Some of them are linked to the rolling stock chosen by the public transport authority: technical innovation can be deceiving. In most of the cases, the profile of the tracks generates a high level of constraints, with many curves and long sections of line at a reduced speed. This is especially the case of the first lines built: they have been designed in order to serve the maximal number of quarters, as if they were the first and also the last line to be built. Changes of direction are frequent and straight lines, along which the rolling stock can run at its nominal speed, are rare. The financing mode, with incitations to serve « disadvantaged » quarters, can explain some diversions. In other cases, the new lines try to correct some past planning mistakes by serving remote service centres like hospitals, universities, etc. The general impression is that urban planning issues (including facade-to-facade reshaping works) are of higher interest than transport ones. .