Summary: A multi-channel queueing system with repeated calls of \(M/M/c/\infty\) type is examined in steady state. A computationally efficient method for computing the blocking probability, mean queue length, etc. is proposed. It consists in approximating the initial queueing system \(S\) by means of an extra “truncated” queueing system \(S_ m\), which differs from \(S\) in that the intensity of repetition becomes equal to infinity as soon as the queue length becomes greater than some level \(m\). The set of statistical equilibrium equations for this queueing system reduces to a finite set of linear equations and therefore may be solved by a computer. If \(m\) is chosen sufficiently large, characteristics of the initial queueing system \(S\) may be calculated with arbitrary accuracy. For example, if the intensity of the arrival process of primary calls \(\lambda =4\), intensity of repetition \(\mu =20\), service intensity \(\nu =1\), the number of channels \(c=5\), \(m=0,1,5,10,20\) then the blocking probability \(B_ m\) in the queue \(S_ m\) is 0.5541, 0.5458, 0.5370, 0.5352, 0.5347, respectively, whereas the exact value of the blocking probability \(B\), i.e. the blocking probability in the queue \(S\), is 0.5347.