Some efficient acylphosphine oxides have been used for many years as photoinitiators of radical polymerization. As very few new acylphosphine oxide compounds have been proposed these past years for the radical polymerization of methacrylates, the aim of this work is to use molecular modeling to select potentially reactive compounds before their synthesis to see whether acyldiphenylphosphine oxide (ADPO) molecules can compete or even work better than commercial and well-established acylphosphine oxides. Two of the common acylphosphine oxides, TPO-L and BAPO, used in industrial applications will be taken as references in this work, while new ADPO will be evaluated in photopolymerization reaction initiated upon LED at 395 nm. In fact, the photopolymerization industrial technology is evolving from the use of an energy-consuming Hg lamp toward the use of safer LEDs working typically in the 385-405 nm range. At least two of the compounds are competing or working better than the well-established systems in a reference benchmarked acrylic resin, evidencing the importance of molecular modeling for the design of new photoinitiators before their synthesis.