The aim of this study was to investigate to what extent population density and small scalehabitat variability affect somatic growth of resident, stream-living brown trout Salmotrutta. There is an ongoing debate whether density-dependent regulation of stream-livingsalmonids operates via mortality and emigration or growth. A capture-mark recapturestudy of brown trout was performed in Bellbekken, a small stream in east Norway. The1.5 km stream was divided into 25 contiguous Sites. Trout movement was limited, and76 % of the recaptures were caught at the same Site as the previous capture. Microhabitatquality and trout density were estimated and used as predictors in the statisticalmodeling of instantaneous growth rate. Growth was estimated from length increments ofrecaptured fish, and from back-calculated lengths, determined from scale readings. Ageneral linear model (GLM) showed that instantaneous growth rate was negativelyrelated to length and age, and that immature fish grew better than mature fish. Afteradjusting for differences in length, age and the maturity state, ~40 % of the variability ingrowth was not accounted for. Some of this variability was related to Site, but most of itremained unexplained. Fish density and estimates of habitat quality accounted forroughly half of the variation explained by Site. The remaining variation could be relatedto other habitat variables which were not estimated, or alternatively that the habitatvariables in this study were not measured at the most relevant spatial scale.