Appropriate modeling of maintenance downtime distribution requires sufficient and good quality data. However, in the absence of those conditions of data, an alternative method using expert opinion inputs is usually employed. Literature on the applications of expert opinion in modeling maintenance downtime is found fairly limited. This study aims to fill in that gap and suggest a practical way of incorporating expert opinion in the modeling. A case study of an offshore gas compression train system is presented here to describe the approach used in modeling of preventive maintenance downtime using expert opinion. Using right set of questionnaires and a fractile technique, downtime distributions are successfully elicited from expert. The comparison analysis shows that the downtime distribution derived from expert opinion is tighter than the one estimated solely based on raw plant maintenance data, indicating that the expert opinion approach is a good alternative for estimating the downtime measures of maintenance tasks. © 2011 Asian Network for Scientific Information.