Growth normal faults propagate syndepositionally at the depositional surface and their dips flatten with depth as the surrounding sediments compact, as described by the equation

<fr>tan ^Thgr</>tan ^Thgr₀</fr> = <fr>(1-^phgr₀)</>(1-^phgr)</fr>,

where ^phgr₀ and ^phgr are the initial and final porosities of the surrounding sediments and ^Thgr₀ and ^Thgr are the initial and final fault dips after compaction. This equation is tested on several southern Louisiana growth normal faults whose shapes are known from well penetrations and whose porosity/depth relationships for sand and shale are derived from well logging. The initial surface fault dip ^Thgr₀ = 67° and the initial sand and shale porosities ^phgr₀^ss = 39.3% and ^phgr₀^sh = 68%, respectively, were determined by an inversion technique on an extensive data set from one fault. Once these initial conditions were determined, the shapes of other faults were computed accurately based on knowledge o the sand-shale stratigraphy from single well logs. The compaction equation allows accurate determination of the proportion of fault flattening caused by compaction and other processes. For example, one fault flattens from an initial dip of 67° to a dip of 38° at 5 km depth; 23° of the flattening is caused by compaction and 6° is caused by footwall folding. Finally, in pure shale the compaction-induced flattening of a growth fault is as much as 19° in the first 500 m of burial and reduces to about 1°/km below this depth; in pure sand the flattening is approximately 1°/km throughout.