In mathematics — specifically, in measure theory — a perfect measure (or, more accurately, a perfect measure space) is one that is "well-behaved" in some sense. Intuitively, a perfect measure μ is one for which, if we consider the pushforward measure on the real line R, then every measurable set is "μ-approximately a Borel set". The notion of perfectness is closely related to tightness of measures: indeed, in metric spaces, tight measures are always perfect.

A measure space (X, Σ, μ) is said to be perfect if, for every Σ-measurable function f : X → R and every A ⊆ R with f⁻¹(A) ∈ Σ, there exist Borel subsets A₁ and A₂ of R such that

${\displaystyle A_{1}\subseteq A\subseteq A_{2}{\mbox{ and }}\mu {\big (}f^{-1}(A_{2}\setminus A_{1}){\big )}=0.}$

• If X is any metric space and μ is an inner regular (or tight) measure on X, then (X, B_X, μ) is a perfect measure space, where B_X denotes the Borel σ-algebra on X.

• Parthasarathy, K. R. (2005). "Chapter 2, Section 4". Probability measures on metric spaces. AMS Chelsea Publishing, Providence, RI. ISBN 0-8218-3889-X. MR 2169627.
• Rodine, R. H. (1966). "Perfect probability measures and regular conditional probabilities". Ann. Math. Statist. 37: 1273–1278.
• Sazonov, V.V. (2001) [1994], "Perfect measure", Encyclopedia of Mathematics, EMS Press