The tidal gravitational field of a cluster distorts the shapes of backgroun
d galaxies. This signature can be used to place constraints on the mass pro
file of the cluster, using a maximum likelihood analysis to find the best-f
it parameterised model. In Schneider et al. (2000) and King & Schneider (20
01) we investigated how accurately parameters can be obtained, concentratin
g on smooth profiles. Here, we consider a realistic model for substructure
in a cluster (from N-body simulations), in order to see what influence this
has on the recovery of parameters. We find that for a typical cluster at z
(d) = 0.2, the presence of substructure increases the error on parameters r
ecovered from weak lensing data by approximate to3%. Even doubling the subs
tructure amplitude induces only a approximate to 10% increase in the parame
ter dispersion: this bodes well for our method.