In design rules for the basal geosynthetic reinforcement (GR) of a piled embankment, the GR strips between each pair of adjacent pile caps are assumed to attract most load. Van Eekelen et al., (2015) and Zaeske (2001) showed that this is a realistic assumption. This paper analysis two aspects of the load distribution on these GR strips: (1) how to distribute the load between the transverse and longitudinal GR strips for non-square rectangular pile patterns and (2) how to further distribute the load on each GR strips between a pair of adjacent piles. In a piled embankment with a non-square rectangular pile pattern, Zaeske (2001) requires least reinforcement in the direction with the largest pile spacing. The Concentric Arches model of Van Eekelen et al. (2013, 2015) requires most reinforcement in that direction with the largest pile spacing. The last one seems a better approach. Van Eekelen et al., (2015) showed that when there is no subsoil support, or almost no subsoil support, the inverse triangular load distribution on the GR strips gives the best match with a large number of measurements. However, this load distribution is a schematization of the real load distribution that is probably more parabola-like and does not have a zero central point. This paper compares the inverse triangular load distribution with these alternative load distributions and concludes that the difference is very limited. Therefore, the relatively simple inverse triangular load distribution is sufficiently adequate for the situation without or nearly no subsoil support.