While green roof (GR) systems have gained popularity as storm water management tools, more emphasis has been applied to studying performance aspects, including stormwater retention. Of particular importance is the substrate layer in which the vegetation grows, which contributes the majority of stormwater retention capabilities. This research investigated many aspects of GR substrate performance, including component durability and component effects on hydraulic conductivity, matric potential, and plant growth. Several commercial substrate blends were tested for durability against successive freeze/thaw cycles with before and after-treatment granulometric distribution analyses. All substrate blends showed significant (p<0.05) particle degradation after 30 freeze-thaw cycles, compared to German (FLL) guidelines. The hydraulic conductivity and matric potential of three experimental GR substrates with increasing volumetric proportions (10%, 20%, 40%) of organic matter (OM), were determined using the HYPROP© method, which extends the traditional measurement range for soils. However, the high porosity of GR substrates resulted in tensiometer water column cavitation near -30kPa. Further studies with the same experimental substrates and OM ratios included both growth chamber studies to rigorously quantify the effects on plant growth and evapotranspiration and outdoor platform experiments to determine effects of OM content on stormwater retention. Growth chamber studies with Sedum kamptschaticum showed that increasing substrate OM increased plant root and shoot biomass. Consecutive periods of water stress showed no differences in evapotranspiration between planted substrate OM treatments levels, but greater water loss was noted from the planted treatments compared to unplanted controls (p<0.05). Substrate volumetric water content (VWC) during the stress periods reached 5% VWC for all planted treatments and all dry-down periods, highlighting differences in plant-available water between these and the laboratory results. While outdoor platform studies showed no effects of OM content on stormwater retention, increasing organic content increased plant canopy coverage (p<0.05). It is likely that differences in retention will be more defined over time as the system matures. Stormwater retention data represented the second growing season for the experimental platforms; given the effects of organic matter on plant growth, analysis of three- or even five-year retention will likely better predict the effects of organic matter on stormwater performance.