1. To gain a better understanding of the heterotrophic nature of small headwater streams in forested landscapes we explored the spatial and temporal variability of in-stream organic matter processes. Three methods were used to measure the benthic metabolism of different in-stream habitats in seven streams throughout a calendar year. This allowed us to analyse the contribution of various metabolic habitats (i.e. sediment, leaf litter, cobbles) to in-stream metabolism during a natural flow regime. Furthermore, it allowed us to define in-stream patchiness based on functional rather than structural elements. 2. Bacterial growth, measured using a leucine assay, displayed a quadratic relationship over time with a peak in warmer months and consistently higher bacterial growth in fine depositional (3.00�710.64 mg C m)2 day)1) than coarse gravel (38.84�582.85 mg C m)2 day)1) sediments. 3. Community metabolism, measured using dissolved oxygen chambers, showed distinct diel patterns and consistently greater net daily metabolism in leaf packs ()261.76 to )24.50 mg C m)2 day)1) than ï¬ne depositional sediments ()155.00 to )15.56 mg C m)2 day)1). Coarse gravel sediments ()49.55 to )16.88 mg C m)2 day)1) and cobble habitats ()151.98 to 55.38 mg C m)2 day)1) exhibited the lowest metabolic rates. Modelled whole-stream metabolism was highly variable among streams and temporal patterns appeared driven by temperature and the relative contribution of patch conï¬guration as a function of flow. 4. Cellulose decomposition potential showed higher rates of microbial activity in fine depositional compared to coarse gravel sediments (30.5 and 29.1 kg average cotton tensile strength loss respectively), though there were higher rates of thread loss indicative of macroinvertebrate activity in gravel compared to depositional sediment (21% and 13% average thread loss respectively), with a slight quadratic trend. The high variability among habitats, streams and over time in this integrative measure may be explained by variability in local microbial activity as well as the potential for macroinvertebrates to contribute across patches. 5. There were strong relationships among benthic processes and habitat structure, nutrient status, stream temperature and flow. Different habitats had distinct metabolic character- istics and these characteristics appear to influence stream food webs and biogeochemical cycling depending on the relative abundance of habitats. Generally, within habitat variability was less than among habitat variability and among stream variability was less than temporal variability. Hence, in terms of the spatial and temporal heterogeneity of benthic processes, these small headwater streams showed predictable metabolic patterns. However, there were few correlations between differing measures of benthic metabolism at the same patch and this suggests that caution should be taken when attempting to infer the rates of one level of metabolic activity (e.g. whole community metabolism) based on another (e.g. bacterial productivity).