The temporal and spatial evolution of thrust fault arrays is currently poorly understood, and marine fold and thrust belts at the toe of passive margin gravitational systems, imaged by commercial 3D seismic reflection datasets, afford a unique opportunity to investigate this problem in three dimensions. Using an extensive 3D seismic data set and age data, the total cumulative strain (shortening) and interval strain rates have been calculated for 11 thrust-related folds mapped in the toe-thrust region of the southern lobe of the Niger Delta. For the first time, the sequence of thrust nucleation, propagation and linkage through time at a scale of 10 s km both along and across strike is documented. Short thrust segments had nucleated throughout the entire study area by 15 Ma. They then grew largely by lateral growth and linkage, increasing the fault trace length and generating asymmetric strain-distance plots, for the first 50% of their history. Thereafter, growth continued by shortening, with minimal along strike increase in fault length. Changes in shortening-distance data between adjacent structures across strike suggest that the change in growth mode occurred once the thrusts had linked in 3D through the common underlying detachment. Over the entire thrust array the strain rate varies through time, starting slowly (<200 m/Ma), then increasing between 9.5 and 3.7 Ma (200–400 m/Ma) before slowing down in the last ∼ 4 Ma (<150 m/Ma). The variation in strain rate is attributed to a change in boundary conditions of the gravitational system. An increase in sediment supply to the delta occurred in the late Miocene-Pliocene, driving higher shortening rates in the toe area. A subsequent reduction in sediment supply in the last ∼4 Ma led to a reduction in deformation rate and the cessation of activity on a number of the thrusts. Predictions of the critical taper wedge model are used to explain the near-synchronous growth of the entire thrust array over the last 15 Ma. Because sedimentation acts to lower the surface slope, the wedge can only continue to deform if shortening occurs over a wide area allowing the surface slope to build up. These new results suggest that models of piggyback fault propagation are not appropriate for deep-water fold and thrust belts.