Recent measurements of trace metals in sediment pore waters at high spatial resolution have revealed significant horizontal and vertical heterogeneity on a submillimeter scale. These measurements are consistent with remobilization occurring from a three-dimensional (3D) stochastic distribution of small “microniche sources”. However, early diagenetic processes are conventionally described in 1D terms. Application of 1D reaction-transport models to 3D systems will result in biased estimates of process rates. For the same intrinsic rates of supply and removal, maxima in concentration-depth profiles in 3D systems are likely to be lower, and concentration profile gradients higher, than in 1D systems. The simple examples considered suggest that process rate estimates may be in error by a factor of 5 when a 1D model is used. A simple 3D numerical model of trace metal remobilization in pore waters was used to demonstrate how the structure of high-resolution trace metal profiles can be reproduced using a stochastic distribution of microniche sources. Heterogeneity depends on the scale considered and is more marked when measurements are made at high resolution. Heterogeneity is increased by slow transport, fast sinks, and widely separated sources. As the degree of heterogeneity between and within concentration-depth profiles increases, the estimates of process rates obtained from 1D models become less accurate.