A measure of complexity for 1/f(beta) fluctuation is proposed using the blo
ck entropy of a sequence of correlated symbols generated by 1/f(beta) fluct
uation. This measure of complexity represents the diversity of sequences of
correlated symbols. It is shown that the measure of complexity has the max
imum value for 1/f fluctuation among 1/f(beta) fluctuations when the length
of sequence of correlated symbols is chosen appropriately.