The authors proceed with the study of recursive fractions started in Part I [Mat. Metody Fiz.-Mekh. Polya 54, No. 1, 57–64 (2011); translation in J. Math. Sci., New York 183, No. 1, 54–64 (2012; Zbl 1274.40009)]. Algorithms are constructed for calculating the value of the expression \(P_kQ_n-P_nQ_k\), where \(\frac{P_k}{Q_k}\) and \(\frac{P_n}{Q_n}\) are correspondingly the \(k\)-th and \(n\)-th rational truncations of a recursive fraction. The values of this expression are used to conclude on the nature and rate of convergence of rational truncations of the recursion fraction to its value.