We evaluate the use of the Bartholomew Sp/Pr index with reference to the collection of information to be used in index calculation and interpretation. Through examination of hypothetical structural arrangements, we address issues related to determining the spacing between structures which are non-uniformly distributed. We also utilize field-collected data to examine how measures of inter-structural spacing may vary when (1) 2 different field methods are used and (2) measurements of the same structures are made over different spatial scales. Using a mangrove prop-root system, measures of inter-structural spacing within 2900 m(2) areas using a 100 point confined arc method were obtained, as were small quadrat samples conducted within the same sites using the confined arc and nearest neighbor method. In small quadrat sampling, mean between-root distances using the arc method were nearly 20% less than those obtained from the 900 m(2) plots. The greatest difference in spacing measurements, and accordingly Sp values, was recorded when the confined arc and nearest neighbor methods were applied to the same inter-structural spaces. Our combined results illustrate that quantification of spatial arrangement can vary with both method and scale of measurement and suggest that across-study comparisons using the index developed by Bartholomew et al. (2000, Mar Ecol Prog Ser 206:45-58) will be difficult to interpret unless the same method of structural complexity measurement is employed, and similar spatial arrangements are compared.