This paper presents findings from framed field experiments on Approval Voting that took place during the 2002 and 2007 French presidential elections. We describe the experimental designs and the main results. Various lessons are drawn from the experiments: (i) Further such experiments are feasible, and the previous ones were very well received by voters; (ii) The principle of approval voting is easily understood and accepted; (iii) Within the observed political context, approval voting modifies the overall ranking of candidates compared to the official first round voting method, and, further, it is shown that different voting rules may yield different outcomes; (iv) lastly, we review various interesting findings that research programs have derived from deeper analysis of the data collected.