Managers of ageing structures aim to extend their service life through timely inspections and preventive maintenance. Currently, planning and decisions on maintenance activities are based on condition assessment methods that rely primarily on visual inspections. However, experience indicates that the deterioration is already in an advanced state at the appearance of the first visual signs of distress and precludes timely preventive maintenance interventions. In addition, the analysis of a sequence of ratings solely based on the evolution of visually based ratings provides very little insight on the rate of the deterioration, and the estimation of residual service life. The objective of this paper is to propose a procedure that supplements visual inspection reports based on structure specific nonlinear simulations of deteriorations associated with the ingress of chloride ions, corrosion of the reinforcing steel, and the cracking and spalling of concrete. Structurespecific simulations are obtained though a model that uses site-specific hourly historical meteorological data, and concrete properties from non-destructive permeability and resistivity measurements. The model is used to estimate the time to the initiation of corrosion as well as the time to the first corrosion-induced cracks to anticipate and evaluate pre-visual conditions. The proposed methodology is demonstrated for an ageing structure in Montreal (Canada). The results show that visual inspections fail in detecting the onset of corrosion for preventive maintenance, and result in prematurely deficient structures. Model simulations suggest that preventive maintenance on the bridge at year 20, corresponding to the bifurcation point of accelerated deterioration rates, would have been optimal.