The interfacial bond between rebar and concrete is a critical factor that influences the performance of reinforced concrete (RC) structures because it affects the effective force transfer between concrete and rebar and hence the composite actions. In finite element modelling of RC structures, the bond-slip relation between rebar and concrete is generally neglected due to the increased model complexity by considering debonding behaviours. Limited studies show the debonding performance is loading rate dependent, neglecting it in numerical modelling may lead to inaccurate structural response predictions, but there is no systematic investigation of bonding performance on responses of structures under load of different loading rates. The influences of neglecting debonding in structural response analysis are therefore not properly defined yet. In this study, the finite element model was established and validated by the available impact test data. The lateral resistance and damage modes of the column considering and neglecting bond-slip behaviour between rebar and concrete were examined under various loading rates ranging from 50 kN/s to 1.0 × 106 kN/s. It is found that neglecting bond-slip relation in finite element modelling of RC columns leads to underprediction of the peak force and damage. The assumption of perfect bond has a significant influence on predicting the peak force and damage of column under low loading rates, but its influence reduces with the increase in loading rate, especially when the loading rate exceeds 1.0 × 105 kN/s. The maximum relative difference in peak force between considering and neglecting bond-slip at 50 kN/s is 18.63 %. Parametric studies were conducted to investigate the effects of bonding performance on columns of different width, depth, height, and concrete strength. Based on the numerical results, an analytical formula is developed to quantitatively predict the errors in peak force predictions by neglecting debonding between concrete and rebars of RC columns. The results highlight the importance of considering bond-slip in finite element modelling of RC structures subjected to quasi-static and low-rate dynamic loadings, and demonstrate it is reasonable to neglect debonding in structural response analysis under high-rate dynamic loadings.