Cost/time estimation in flat plate processing (FPP) is a complex system of man-work-environment-organization. As the share and impact of physical, psychological, personal, and other characteristics of human in this complex system becomes bigger, the state of FPP time estimation changes from well defined to ill-defined (uncertain). Due to so many inter-related, imprecise, and ill-defined factors involved in activity time, it is too complicated; time consuming and expensive to model these factors into a universally accepted activity time/cost estimation system using mathematical or even analytical-experimental approaches. This paper addresses the issue of uncertainty treatment in activity time estimation in FPP. The emphasis is placed on modeling of nonrandom uncertainties using fuzzy sets.