AbstractDue to climate change, crop production will increasingly be affected by water limitation, causing remarkable decreases in grain yields of cereals. Plant growth regulators such as paclobutrazol (PAC) have been shown to protect plants from detrimental impacts of drought stress, and improvement of root growth and antioxidant activity were identified as main reasons for their positive effect. A container experiment was conducted with two maize (Zea mays L.) cultivars, Galactus and Fabregas, to investigate how PAC application affects root growth and grain yield under stress conditions. At growth stage V8, the plants were treated once with PAC (0, 2, or 3 mg PAC per plant), and concomitantly reduction in soil water content commenced until 30–35% of the maximum water‐holding capacity (WHC) was achieved. The plants were exposed to this drought condition for three weeks during flowering as the critical period for kernel setting. Both factors, PAC application and drought stress, caused decreases in plant height, whereas total leaf area was unchanged and transpiration rate was significantly reduced by water limitation only. Flowering was almost unaffected by PAC treatment; yet, drought stress significantly delayed start of silking. The straw yield was decreased due to PAC and drought stress, and an improvement of the harvest index was obtained for drought‐stressed Galactus plants with PAC application. Grain yield was unaffected by PAC application, whereas drought stress caused significant decreases by 15% on average of both cultivars. The kernel number of drought‐stressed Galactus plants was increased after PAC treatment, but concurrently smaller kernels were produced. Water limitation generally decreased kernel number. Drought‐stressed Fabregas plants consumed less water after PAC treatment, resulting in significant improvements of water‐use efficiency (WUEgrain) during silking and thus most likely alleviating stress intensity. For both cultivars, PAC treatment and water limitation showed almost no significant impact on root dry matter, root length density, and root surface area, either determined for different soil layers down to 80 cm or on a per‐plant basis. It is concluded that grain yield performance of maize plants, exposed to water limitation during flowering, was not source‐limited but sink‐limited. Consequently, even if PAC can cause improvement of antioxidant activity and photosynthesis, due to sufficient availability of assimilates in the maize kernels a positive effect on grain yield is improbable. Considering source–sink relationships during flowering and kernel set, enhanced root growth due to PAC treatment did apparently not occur.