Free-form surfaces are widely used in engineering applications. These surfaces are complex and without rotational symmetry, and for this reason they are inspected using the coordinate measuring machines equipped with contact sensor require a suitable sampling strategy. Sampling algorithms are one of the most important factors of error creation in the accuracy of substitute geometry. In coordinate measuring machines, the sampling strategy involves the estimation of the number of sample points (sample size) and identification of their positions (how distribution) on the surface. Thus sample points should be distributed on the surface using sampling strategies that are appropriate for the surface. Often it is difficult to establish such pieces of information (number and the way of distributing the points on the surface) owing to the complex nature of free-form surfaces. In the present work for first time, new adaptive sampling strategy by particle swarm optimization algorithm (PSO) for sampling from free-form surface is proposed. The proposed strategy was compared with two conventional strategies and the deviation between substitute geometry and CAD model is extracted. The simulation results showed that in the proposed method the deviation between substitute geometry and CAD model is less than conventional methods by 2 to 3 times (depending on the number of points). Therefore high efficiency of the proposed method over other methods is concluded.