We consider, within a Brownian motion framework, estimation of secondary pa
rameters following a sequential test. Whitehead (1986b) proposed a method f
or reducing the bias of maximum likelihood estimators. In this paper we pre
sent unbiased estimators to eliminate the bias.