In this work, it is presented a theoretical-experimental study of acoustic perturbations propagation, in turbulent, subsonic, air flow in a circular pipe (lenght: 3000mmm; diameter: 50,8mm). Sinusoidal perturbations were introduced in the flow, by means of a loud-speaker, placed at the wall of a settling chamber, upstream of the test pipe. The wave propagation along the flow was studied, as well as its effects on pressure, velocity and turbulance intensity distributions. The experiments were conducted at flow Reynolds Number 70000, introducing acoustic perturbations corresponding to the first and second resonant frequencies of the pipe, namely 56Hz and 112Hz (Strouhal Numbers 0.13 and 0.26). At 56Hz the intensity of the acoustic perturbation was 3 percent (r.m.s. value of the wave component of the velocity, at the perturbation frequency, at the pipe entrance, normalized by the pipe entrance centerline mean velocity); at 112Hz two perturbation intensities were applied: 3 percent and 18 percent. The mean velocity, turbulance intensity and wave component profiles were measured at several stations along the pipe, between r/R=0 and r/R=0,96. The wave propagation along the pipe was theoretically studied. Two models were considered, a model without dissipation and a model with dissipation. The experimental results confirmed the dissipation estimates based on the model, and have shown that, for the studied case conditions (namely for the frequency range considered, and pipe lenght of the order of the wave lenght), the dissipation has a moderate effect on the wave propagation. So, a great part of the wave behavior is a interpreted on the basis of the model without dissipation, that has shown good agreement with the experimental results. The model with dissipation allows to intepret some aspects strongly connected with dissipation, namely the wave behavior in the vicinity of the ressonance and the transversal profiles of the wave component of the velocity, in the studied case, of the model without dissipation. The differences between the behavior forseen by the presented models, and the experimental results, were of the order of magnitude of the measurement errors. According to the analysis performed, it appears that, for the studied conditions, neither the turbulance structure significantly affects the acoustic wave, nor the acoustic perturbations significantly affect the turbulent flow characteristics (mean velocity, turbulance intensity and pressura distribution). So, it appears that, a linear approach, based on a superposition technique, used in the presented theoretical models, is adequate to describe the overall disturbed turbulent flow.