Abstract: Direct computation of the weighted sum of N complementary error functions at M points scales as O(MN). We present a O(M + N) $\epsilon$-exact approximation algorithm to compute the same.