The objective of this study was to identify biomechanical predictors for accuracy and speed of the wrist shot in floorball, comparing two different starting feet positions. Ten floorball players performed 2 series of 10 stationary wrist shots, in 2 different positions (feet at a right angle to the end of the stick, oriented towards a target and feet parallel to the end of the stick and to the target). A 12-camera motion capture system, tracking reflective markers on key landmarks, was used to record participant and stick kinematics. Accuracy of the shot was quantified by distance of impact from target centre. Player gaze was approximated from head position. Shot accuracy was significantly better (P = 0.007) when feet were at right angle (0.22 [0.14] m) than when they were parallel (0.27 [0.20] m). Ball speed was no significantly different (P = 0.485) between the right angle position (23.50 [17.52] m · s(-1)) and the parallel position (23.50 [17.95] m · s(-1)). Between self-selected position and imposed position, there was no significant difference. Players looking at the target during shooting had greater accuracy. Regression models suggested that ball speed was mainly influenced, in both positions, by the flexion of the supporting leg (ankle, knee and hip), by the rotation of the hip and of the trunk, especially for the spine angles, and by the rotation and abduction-adduction movements of the wrist of the hand on the top of the stick. The comparison between players showed important differences in these technical skills.