The objectives of this research were to assess the physical properties of six different growing substrate mixtures destined for roof gardens and determine the influence of these substrates on the morphology, physiology, growth and flower quality of pansy (Viola × wittrockiana), Madagascar periwinkle (Catharanthus roseus), and Pavia lily (Longiflorum×Asiatic lilies (Lilium) ‘Pavia’). Six green roof growing substrate mixtures (by volume) were used [T1: coarse tuff+fine tuff+cocopeat (5:1:4), T2:coarse tuff+medium tuff+fine tuff+cocopeat (5:5:2:8), T3: medium tuff+fine tuff+cocopeat (5:1:4), T4:medium tuff+fine tuff + peat moss (5:1:4), T5: perlite+medium tuff+fine tuff+cocopeat (5:5:2:8), and T6: lightweight expandable clay aggregates+fine tuff+cocopeat (5:1:4)]. The T6 (LECA-cocopeat) had the lowest weight at field capacity and good aeration and WHC to sustain optimal plant growth. Medium tuff-peat moss (T4) produced the highest pansy flowers number per plants. However, peat moss (T4) has been identified by environmentalists as an unsustainable media. LECA-cocopeat (T6) had a higher number of lily flowers and leaf area than T1-T4. The cost of the LECA substrate ($US 215 m−3, T6) is extremely higher than that of volcanic tuff ($US 36 m−3, T1-T4), and perlite ($US 100 m−3, T5). Overall, T6 can be an ideal option in terms of physical properties of growing substrate and flower quality but it might not be the best option for green roofs when cost is the primary concern. In addition, certain growing substrates could be used to satisfy specific growing requirements without sacrificing performance. For example, the combination of medium tuff and peat moss (T4) produced the highest pansy flower numbers per plant but required a longer period to flower (95 days). So, this growing substrate could be used where plentiful blooms are desired and the time to bloom is not a concern.