Seismic refraction and reflection, gravity and well data have been used to constrain the structure and evolution of the Valencia trough: a young extensional basin in the western Mediterranean. Seismic refraction data reveal the present-day crustal structure, while backstripping of seismic reflection and well data reveal the structure at the time of rifting. By comparing the seismically constrained crustal structure to that inferred from backstripping, we have been able to evaluate whether the Moho is in the expected position, given the tectonic subsidence/uplift and sediment loading history of the trough. The best overall fit between the predicted and seismically constrained crustal structure in the trough is for a model in which the extended continental crust and lithosphere has acquired little or no strength following rifting. A low strength is in accord with the observed free-air gravity anomaly as well as results from other types of rift basins such as the North Sea and the continental margin off the east coast of the USA. There are discrepancies between the observed and predicted Moho depths beneath the flanks of the Valencia trough, but they are attributed to either the removal of crustal material during or following rifting (Iberian peninsula), or to orogenic loading of previously extended continental crust (Mallorca). The largest discrepancy (up to 10 km) correlates with the Ebro delta which has prograded into the trough by more than 60 km since the Pliocene. The discrepancy cannot be explained by the same large-scale processes that modified the Moho beneath the trough flanks and requires that the lithosphere beneath the Ebro delta region is associated with a high (60 < Tc < 80 km), rather than low, flexural strength. We do not believe, however, that the high values indicate that extended continental lithosphere is fundamentally strong. Rather, we attribute it to incomplete stress relaxation of the lithosphere during loading.