Abstract Purpose The force–velocity relationship of muscular contraction has been extensively studied. However, previous research has focussed either on isolated muscle or single-joint movements, whereas human movement consists of multi-joint movements (e.g. squatting). Therefore, the purpose of this study was to investigate the force–velocity relationship of isovelocity squatting. Methods Fifteen male participants (24 ± 2 years, 79.8 ± 9.1 kg, 177.5 ± 6 cm) performed isovelocity squats on a novel motorised isovelocity device (Kineo Training System) at three concentric (0.25, 0.5, and 0.75 m s⁻¹) and three eccentric velocities (− 0.25, − 0.5, and − 0.75 m s⁻¹). Peak vertical ground reaction forces, that occurred during the isovelocity phase, were collected using dual force plates (2000 Hz) (Kistler, Switzerland). Results The group mean squat force–velocity profile conformed to the typical in vivo profile, with peak vertical ground reaction forces during eccentric squatting being 9.5 ± 19% greater than isometric (P = 0.037), and occurring between − 0.5 and − 0.75 m s⁻¹. However, large inter-participant variability was identified (0.84–1.62 × isometric force), with some participants being unable to produce eccentric forces greater than isometric. Sub-group analyses could not identify differences between individuals who could/could not produce eccentric forces above isometric, although those who could not tended to be taller. Conclusions These finding suggest that variability exists between participants in the ability to generate maximum eccentric forces during squatting, and the magnitude of eccentric increase above isometric cannot be predicted solely based on a concentric assessment. Therefore, an assessment of eccentric capabilities may be required prior to prescribing eccentric-specific resistance training.