- It is generally accepted in stream ecology that habitat heterogeneity and patchiness at multiple scales increases ecosystem resilience through niche diversification. Heterogeneous stream habitats include a complex mosaic of hydraulic features, large woody debris, anabranches, substrata and channel forms - this complexity tends to increase as streams progress towards later stages based on the Stream Evolution Model. Recent restoration work on Whychus Creek in Central Oregon has sought to create complex, late-stage systems in order to improve the ecological function of artificially simplified reaches. One way to measure and track lotic system habitat complexity is through substrate analyses because a heterogeneous patchwork of substrata can act as a proxy for diversity of aquatic habitat types. The goal of this pilot project was to develop a replicable and robust monitoring protocol that quantifies substrate heterogeneity conditions among four priority reaches in Whychus Creek. To do this, I developed a monitoring protocol that utilizes three methods to capture substrate heterogeneity on four, 500-m reaches of the creek. Each sample reach included a nested sampling design of 12 floodplain-wide transects that allowed me to quantify micro, meso or macro-level substrate heterogeneity. I collected data using standard pebble counts, two-dimensional areal plot estimates and one-dimensional patch width measurements. I used the data from each of these three methods to calculate habitat heterogeneity using four metrics – Simpson’s Diversity Index, Shannon’s Evenness Index, Lloyd’s Index of Patchiness and Fortin’s Spatial Diversity Index. The results indicated that the two recently restored reaches were on average, 38% more heterogeneous than the untreated reach while the older, more established project reach was on average, only 15% more heterogeneous. The chi-square test for independence for the pebble count indicated significant differences between all the reaches and substrate classes (X2 (18, N = 1865) = 210.23, p <.001) except one – which signaled that the untreated reach requires a slightly larger sample size in future years. For the plot method, the differences among the reaches were more significant with X2 (18, N = 2306) = 836.57, p <.001. The plot method resulted in the highest Cramer’s V value of 0.35 (p < 0.001) – indicating a strong relationship between substrate composition and individual reach. These results illustrate that the three methods were robust enough represent stream substrate conditions. Recommendations include, foregoing the pebble count in order to prioritize larger sample sizes of the plot method and the transect patch method.