Workload can be regarded as the “cost” (in information processing terms) of performing a given flight task. Predictive models of workload are valuable at the early stages of system design to identify potentially excessive demands on the human operator before major system development commences. This paper describes the early development of a task-based predictive workload method. It utilizes a scenario-based timeline analysis supplemented by characterizations of the cognitive nature of the pilot’s tasks using Wickens’ multiple resource theory, followed by the identification of likely task conflicts and estimates of the influence of the task environment. Workload predictions were made for a manual approach, landing and roll-out scenario in a modern airliner. Predictions were compared against corresponding Bedford scale workload estimates. Results show that the predictive workload technique produced comparable estimates of cognitive workload measured using the Bedford scale.