As ship bridge design is supposed to support the watch-officer's sensing, information processing and motor-handling activities, the question is raised as to how these processes can be measured. The suggestion is to conduct experiments on different paradigms which together allow general conclusions and interpretation. Two such experiments are described. In the first the effects of two bridge designs on ship control are compared in a simulator experiment with realistic conditions. In a second experiment the effects of radar and view are investigated in a ship tracking task with forcing functions based on the inherent controllability of the ship. The results show that suboptimal bridge design has the effect of deteriorating mariner's task performance which parallels the findings of accident analysis. Future research on ship bridge design is focussed on one-officer-operation.