Effect of different concentrations of paclobutrazol and uniconazole supplied to the recycling nutrient solution in an ebb and flow system on the growth and flowering of Kalanchoe 'Rako' was examined. The cuttings of K. blossfeldiana 'Rako' were planted in 72 cell plug trays containing different volume ratios of cellular glass foam (CGF), granular rockwool (GR), peatmoss (PT), and perlite (PL). The cuttings were rooted in a mist (10 sec every 10 min) propagation bed with a mean temperature of 18.2ºC and RH of 66.7% under a long day regime. Rooting in GCF or PL alone was poorer than the other treatments. But, rooting increased when incorporation ratio of GR and PT was increased. Plants were potted in 10 cm (370 ml) pots and supplied with a nutrient solution containing 80 ml per pot of 0.5, 1.0, or 2.0 mg l^-1 paclobutrazol or of 0.125, 0.25, or 0.5 mg l^-1 uniconazole. For recycled subirrigational supply, plant growth retardants (PGRs) were treated five times at every five days during the short-day condition. The remaining solution was collected in a nutrient solution tank and recycled. Foliar spray was conducted as control for evaluation of the efficacy of subirrigational supply. A foliar spray of 240 ml m² (8 ml/pot) of paclobutrazol 10.0 mg l^-1 or uniconazole 2.5 mg l^-1 was applied twice after pinching at one week interval. Suppression of the plant size, flower stem length and stem diameter were observed from all PGR treatments, compared with the untreated control. However, the number of florets increased in all PGR treatments, compared with the control. Leaf area was reduced less in subirrigational supply of uniconazole than in foliar spray. The treatment of paclobutrazol 1.0 and 2.0 mg l^-1 or uniconazole 0.5 mg l^-1 tended to drop the flowering rate. In the present study, subirrigational supply of paclobutrazol 0.5 mg l^-1 and uniconazole 0.125 mg l^-1 provided the best condition for effectively suppressing stretchiness without deteriorating other important horticultural values in Kalanchoe, 'Rako'. Results of this experiment indicated that subirrigational application of growth retardants at lower concentrations could be used as an effective mean for suppressing stem growth.