The conditional maximum‐likelihood estimates for a binomial logistic latent trait model are derived and examined. A computing algorithm is used to make a comparison between the conditional estimates and a set of unconditional estimates. The general conclusion is that there may be data sets for which significant discrepancies will occur between conditional and unconditional approaches to the estimation of the item parameter