This paper considers the problem of determining the number of nonparam
etric factors in a multivariate nonparametric relationship. The defini
tion given is broad enough to encompass a number of potential applicat
ions in econometrics including inferring the rank of demand, testing w
hether for a given set of instruments it is possible to identify a lin
ear model, and testing arbitrage pricing theory. The paper gives metho
ds for testing hypotheses concerning the number of factors, using seri
es and kernel based nonparametric methods and also considers the probl
em of estimating the number of factors. The methods are compared in a
small simulation study and an application to determining the rank of d
emand systems is considered.