Often simple metrics are used to summarise complex patterns in stream benthic ecology, thus it is important to understand how well these metrics can explain the finer-scale underlying environmental variation often hidden by coarser-scale influences. I sampled 47 relatively pristine streams in the central North Island of New Zealand in 2007 and (1) evaluated the local-scale drivers of macroinvertebrate community structure as well as both diversity and biomonitoring metrics in this unmodified landscape, and (2) assessed whether these drivers were similar for commonly used univariate metrics andmultivariate structure. The drivers of community metrics andmultivariate structure were largely similar, with %canopy cover and resource supply metrics the most commonly identified environmental drivers in these pristine streams. For an area with little to no anthropogenic influence, substantial variation was explained in the macroinvertebrate community (up to 70 % on the first two components of a partial least squares regression), with both uni- andmultivariate approaches. This research highlights two important points: (1) the importance of considering natural underlying environmental variation when assessing the response to coarse environmental gradients, and (2) the importance of considering canopy cover presence when assessing the impact of stressors on stream macroinvertebrate communities.