The issue of morphodynamic influence in meandering streams is investigated through a series of laboratory experiments on curved and straight flumes. Both qualitative and quantitative observations confirm the suitability of the recent theoretical developments (Zolezzi & Seminara 2001) that indicate the occurrence of two distinct regimes of morphodynamic influence, depending on the value of the width ratio of the channel $\beta$. The threshold value $\beta_R$ separating the upstream from the downstream influence regimes coincides with the resonant value discovered by Blondeaux & Seminara (1985). Indeed it is observed that upstream influence may occur only in relatively wide channels, while narrower streams are dominated by downstream influence. A series of experiments has been carried out in order to check the above theoretical predictions and show, for the first time, evidence of the occurrence of upstream overdeepening. Two different sets of experiments have been designed where a discontinuity in channel geometry was present such that the channel morphodynamics was influenced in the upstream direction under super-resonant conditions ($\beta{>}\beta_R$) and in the downstream direction under sub-resonant conditions ($\beta{<}\beta_R$). Experimental results give qualitative and quantitative support to the theoretical predictions and allow us to clarify the limits of the linear analysis.