This paper contributes to the study of the free additive convolution of probability measures. It shows that under some conditions, if measures $\mu_i$ and $\nu_i, i=1,2$, are close to each other in terms of the L\'{e}vy metric and if the free convolution $\mu_1\boxplus\mu_2$ is sufficiently smooth, then $\nu_1\boxplus\nu_2$ is absolutely continuous, and the densities of measures $\nu_1\boxplus\nu_2$ and $\mu_1\boxplus\mu_2$ are close to each other. In particular, convergence in distribution $\mu_1^{(n)}\rightarrow \mu_1,$ $\mu_2^{(n)}\rightarrow\mu_2$ implies that the density of $\mu_1^{(n)}\boxplus\mu_2^{(n)}$ is defined for all sufficiently large $n$ and converges to the density of $\mu_1\boxplus\mu_2$. Some applications are provided, including: (i) a new proof of the local version of the free central limit theorem, and (ii) new local limit theorems for sums of free projections, for sums of $\boxplus$-stable random variables and for eigenvalues of a sum of two $N$-by-$N$ random matrices.