After the studies of Kerr (1987; 1993) who showed that adding a third dimension to imagery facilitates the imagery processes, the study of mental imagery in three-dimensional (vs. two dimensional) patterns has been neglected. The research reported here, comprising three experiments, explores the variables involved in the task of imaging a pathway along a matrix, such as the statement length and the number of to be-remembered positions. Results revealed that the third dimension facilitates the imagery task when participants have to remember the final position (Experiment 1) or the two final positions (Experiment 2) but not if they have to remember not only the last position, but also the complete pathway (Experiment 3). In general, data show that the limitations on the imagery processing are not due to the number of units in each dimension, but to the number of dimensions in which a pathway might move, to the test requirements and to the pathways length. It is hypothesized that the advantage of three-dimensional matrices is due to a better chance of updating the position along the pathway.