Our work has argued for a particular scaling form governing the distribution M(ξ,t) of magnetization over bias ξ, for a system of dipolar-interacting molecular spins. This form, which was found in Monte Carlo (MC) simulations, leads inevitably to a short-time form ∼t1∕2 for the magnetization relaxation in the system. The authors of the Comment argue that the magnetization should decay rather as ∼tp, with the exponent p depending on the lattice type—and they argue this form is valid up to infinite times. They also claim that our conclusion is based on an assumed exponential dependence of the function M(ξ,t) on τde(ξ), the effective molecular relaxation time. In fact, our results do not depend on any such dependence, which was used merely for illustrative purposes, but only on the scaling form we found. Repeating our MC simulations for different lattice types and different parameters, we always find a square root relaxation for short times. We find that the results of the Comment are flawed because they try to fit their results over far too large a range of times (including the infinite time limit, where no simple theory applies).