The upper limb force–velocity relationship (FVR) is a crucial aspect of athletic performance, particularly in para-sports where upper limb movements play a leading role in activities such as wheelchair propulsion. Athletes’ mechanical capacities can be evaluated on the field or in lab conditions. However, no studies have yet indicated a relationship between ergometers and field FVR or performances. Understanding para-athletes’ upper limb FVR can provide important insights for developing effective training programs and improving athletic performance in wheelchair basketball players. Twenty-three wheelchair basketball players (12 women and 11 men) from French national teams performed a battery of three tests consisting of a 20 m sprint (SP) with Inertial measurement units (IMU) on wheels, horizontal upper limb push-offs on a frictionless sled with two Kistler force plates (HBP), and a crank ergometer sprint test (CES) derived from arm ergometer sprints using Brachumera, Lode. For the SP test, the FVR was computed with an estimation of force from the acceleration, the rolling resistance, and an estimation of the air resistance. Correlations between each variable measured were computed via Pearson correlations in R, assuming a strong relationship when r = 0.7–1.0, moderate when r = 0.40–0.69, and weak when r < 0.1–0.39. Significant differences were assumed when p < 0.05. Strong correlations were found between the results of the three tests conducted. The correlation coefficient between maximal theoretical force (F0) data ranged from 0.85 to 0.88, and that between maximal theoretical power (P0) data ranged from 0.87 to 0.94. However, for maximal theoretical velocity (V0) data, the correlations between the three tests were less important. The performance variables showed a strong correlation with power measured in the HBP test without load. There was a significant relationship between the HBP, CES, and SP variables. There was a correlation between performance on the three tests in our population, especially for power. However, the V0 value of HBP was not representative of any performances. This work uses multiple protocols to assess para-athletes’ performance and shows that upper limb symmetry depends on the gesture of the task and the personal sports’ wheelchair.