Math is certainly required to understand the theory and to synthesize experimental observations into a working model, but I wouldn't put it at more than 50-60%. Most of the math is fairly basic relatively speaking, even at the highest levels of physics.