New Guinea impatiens (Impatiens hawkeri Bull.) `Illusion' were grown in a recirculating subirrigation system under various rates and placements of 14N-6.1P-11.6K (Osmocote; Scotts-Sierra, Marysville, Ohio) resin-coated, controlled-release fertilizer (CRF). Four CRF placements (incorporated, top-dressed, bottom, and dibble) were tested. Incorporated placement yielded slightly greater dry weights than the other placements. A rate experiment tested incorporating from 0.5 to 2 times the fertilizer manufacturer's recommended rate of 7.11 kg·m-3. All shoot growth parameters (height, leaf number, shoot, and root fresh and dry weight) exhibited a significant quadratic response, as exemplified by shoot dry weight, where shoot dry weight increased up to the 1.5× rate, after which shoot dry weight decreased. A quadratic response surface model revealed that the optimum rate response ranged from 1.16× rate for height to 1.47× rate for shoot dry weight. The lower bound of the 95% confidence interval (CI) would be the lowest rate at which one could expect maximum growth response. The lower bound of the 95% CI varied from 0.56× rate for height to 1.30× rate for shoot dry weight. Thus, the lowest rate that would be within the 95% CI for all growth parameters, and thus yield maximum growth response, would be the 1.30× rate. Electrical conductivity (EC) of the growing media increased significantly with increasing CRF rate. At all rates, EC was significantly greater in the top layer than in the middle and bottom layers. Only in the 1.75× and 2× rates did EC exceed the recommended EC levels in the middle and bottom layer. All rates >0.75× exceeded recommended EC in the top layer. Release characteristics and total nutrient balance of the CRF was compared in subirrigated and top-watered systems. There was no significant difference between top-watered and subirrigated treatments for the amount of K recovered in plant tops and released from prills. By day 84, in subirrigation, 46% of the K was still in the prills, 41% was recovered in the plant tops, and 22% was recovered in the medium. Similar results were obtained in the top-watering treatment, except that a lesser amount was recovered in the medium (9%) and a small amount (4%) was recovered in the leachate. The uptake of K by plants and release of K by the CRF were inversely proportional and linear with respect to time. Of the K released from the prills, 77% and 83% were recovered in the plant tops for subirrigation and top-watering, respectively, indicating very high fertilizer use efficiency.