This thesis investigates the mechanisms behind the control of a typical two line kite, where the lines are attached to the kite side by side. This arrangement gives the kite flyer the ability to apply a roll angle to the kite, which then results in a yawing motion. The reason for this yaw rotation has not been adequately described previously. The definitions of roll and yaw for a kite have been re-defined to match the real world behaviour of the kite-bridle-line system. Specifically, these are defined as rotations relative to the lines rather than the kite itself. This detail has been neglected in previous research, and has a significant effect on the turning behaviour of a kite. A static model of a kite represented by flat disks was created. This model allows the out of balance forces and moments to be found for a kite when it is held at any position. When the kite is held with a roll angle applied, the disk angles of attack become unequal. This causes a change in the magnitude, direction, and point of action of the aerodynamic forces on each disk, which can lead to a yaw moment. While this does not give a complete picture of how a kite turns, it does explain one of the two mechanisms that cause a kite to begin to yaw when a roll angle is applied. The other mechanism is due to the velocity of the roll rotation, and is out of the scope of this thesis since a dynamic analysis would be required. The static model showed that any variation to kite geometry or any parameter that affects the equilibrium position of the kite will affect turning response. The most important of these parameters for a simple kite represented by two disks is the dihedral angle. A minimum negative dihedral angle (or anhedral) is required for a kite to turn in the expected direction when a roll angle is applied. The value of the minimum anhedral angle required for this behaviour varies with other parameters, but is generally between 8° and 20°. Other parameters such as bridle geometry also affect the turning response of a kite, primarily because they alter the equilibrium positions of the kite and line. Altering these equilibrium positions has a strong effect on turning response, since it changes the initial disk angles of attack. Additionally, if the kite and line are not aligned perpendicular to each other (which is a rare condition for a kite) a roll angle further changes the disk angles of attack, since the roll angle is applied about an axis relative to the line rather than the kite. An investigation into the effect of varying wind velocity on turning response showed that it has an important effect. Some kites will reverse their response to a given roll angle at some wind velocities, which could make the kite very difficult to control. Additionally, some kites can alter their equilibrium positions sharply with wind velocity, again causing varying turning behaviour as the wind conditions change. Future work should examine the dynamic turning response of kites. A dynamic simulation could be used to examine how the turning response of a kite is influenced by the rate at which a control input is applied. Additionally, the behaviour of the kite once the initial turning movement has begun could be assessed.