Statistical methods for mapping quantitative trait loci relative to ge
netic markers are now well established for continuous traits with norm
al distributions. However, many traits of economic importance are reco
rded on a discrete, ordinal scale. Here we describe a model developed
for the analysis of ordinal traits, such as degree of difficulty in ca
lving or categories of plant disease resistance. The model estimates t
he distance from the quantitative trait locus to neighbouring genetic
markers, and also genetic parameters, either as gene effects on an und
erlying continuous scale or as probabilities of the observed categorie
s. The model is tested on simulated data and is compared with an analy
sis based on mixtures of normal distributions. The ordinal model is fo
und to estimate the parameters more accurately, especially when the nu
mber of categories is small or when only one linked marker is availabl
e.