A common approach to modelling spatial choice is to apply the theory of constrained utility maximising behaviour by individual choice makers within a probabilistic framework. Decision making is assumed to be deterministic but utilities stochastic in the random utility version of the approach. For operational models, parametric distributions have to be specified for the stochastic component of the utility functions. It is commonly assumed in deriving spatial choice models that each of the choice alternatives are perceived by individuals as unique, independent opportunities. This paper considers the case where aggregations of choice alternatives are perceived similarly. The logit distribution (as an approximation to the Gauss (normal) distribution) is chosen as the parametric stochastic utility distribution. The choice model is derived and its properties and potential application discussed.