This study was carried out using two sets of numerical weather forecast data and flight reports for Clear Air Turbulence (CAT) over Iranian Plato to find atmospheric flow patterns favorable to the formation of CAT. The numerical data include five months of AVN analysis with horizontal resolution of 1 degree(about 100 km) and four months forecast data of MM5 model with resolution of 50 km. Important indices associated with conditions favorable for CAT formation include wind shear, flow deformation fields and atmospheric static stability (or joint effects such as Richardson number). These indices were estimated and also some algorithms such as Dutton to find their threshold values for CAT occurrence. The best selected algorithms are determined in the 5 months period using AVN outputs. In another 4 months period the MM5 outputs were used for the forecast of CAT using the POD method. The results for there periods indicated that severe CAT rarely occurred over Iran and condition for moderate CAT are often in upstream of troughs ridges, near straight jet stream with cold advection and upper region of ridges. The POD method shows Dutton and Ellrod-2 are the best selected algorithms for CAT forecasting and have good consistency with observations. Their typical threshold values (associated with wind shears) are about 3 and 20 for moderated CAT respectively.