Tailored femtosecond laser pulses from a computer-controlled pulse sha
per were used to optimize the branching ratios of different organometa
llic photodissociation reaction channels. The optimization procedure i
s based on the feedback from reaction product quantities in a Learning
evolutionary algorithm that iteratively improves the phase of the app
lied femtosecond Laser pulse. In the case of CpFe(CO)(2)Cl, it is show
n that two different bond-cleaving reactions can be selected, resultin
g in chemically different products. At Least in this case, the method
works automatically and finds optimal solutions without previous knowl
edge of the molecular system and the experimental environment.