The central limit theorem for random measures derived from a stationary Poisson point process of compact subsets of the Euclidean space is formulated. Various intensity estimators are compared and approximate confidence intervals are constructed.