It is stated by Ford and O'Connell in this festschrift issue and elsewhere
that "there is no quantum regression theorem'' although Lax "obtained a for
mula for correlation in a driven quantum system that has come to be called
the quantum regression theorem". This produces a puzzle: "How can it be tha
t a non-existent theorem gives correct results?" Clarification will be prov
ided in this paper by a description of the Lax procedure, with a quantitati
ve estimate of the error for a damped harmonic oscillator based on expressi
ons published in the 1960's. (C) 2000 Elsevier Science B.V. All rights rese
rved.