Function approximation and interpolation are major and important problems in various scientific fields. In this work we present an approximation method of fuzzy data defined at a 3D fuzzy data set. We define a fuzzy smoothing bicubic spline approximation for a given fuzzy data set and we estimate the approximation error using similarity measures of fuzzy numbers. Examples are given to test the goodness of the method and compare the behavior of the indices proposed for different configurations of the fuzzy smoothing bicubic spline. Finally, some conclusions of the presented method are briefly discussed.