The approximate conditional likelihood method proposed by \textit{D. R. Cox} and \textit{N. Reid}, J. R. Stat. Soc., Ser. B 49, 1-39 (1987; Zbl 0616.62006) is applied to the estimation of a scalar parameter \(\theta\), in the presence of nuisance parameters. The estimating function of \(\theta\) based on the approximate conditional likelihood is shown to be preferable to that based on the profile likelihood. A sufficient condition for both approaches to be equivalent is given. The role of parameter orthogonality is emphasized. Several examples including bivariate normal means with known coefficient of variation are presented.