This paper reviews intercropping research for a particular case: cassa
va-based crop combinations. Cassava is dominated in combination with m
aize while it is the dominating species in combination with low-growin
g species. Combinations with maize or legumes show a real biological a
dvantage over the sole crops reflected in (modified) Area x Time Equiv
alency ratio (ATER) values above unity. This is not the case with swee
t potatoes. Success of maize+cassava mixtures depends on time and rate
of recovery of cassava after maize harvest. Biological advantage tend
s to disappear when maize yield exceeds about 3.5 t ha-1. Undergrowing
conditions or practices which result in high maize yield, intercroppi
ng cassava with maize is not biologically advantageous. Biological adv
antage of intercropping with legumes decreases with the legumes' growt
h duration, which should not exceed 90 days. Physiological traits of c
assava for successful intercropping with maize or with legumes are pro
bably not the same, but their nature is not clear. Moderate early vigo
ur and a high partitioning of dry matter to the storage roots after ha
rvest of the associated crops seem important in both cases. Cassava br
eeding for sole cropping has resulted in varieties with good performan
ce in intercropping. Whether varieties can be selected with better ada
ptation to intercropping cannot be concluded from the literature. Dry-
matter distribution, in particular after harvest of the associated cro
p, seems important but more growth analytical studies are required. Us
efulness of currently available crop models in the study of intercropp
ing is doubtful.