Upon occasion it is required to find the drag of aircraft components which are normally retracted in level flight, such as flaps and undercarriages. The usual method of obtaining this drag from flight tests, is to measure speeds in level flight under known engine conditions with the component retracted and then with the component down. The difference of speed so obtained, together with the known test conditions then yields an answer. The disadvantage of this method is that, as the difference of speed is usually large, incidence changes and airscrew and engine assumptions cause discrepancies in the results. The difference of power between different engines of a type can be ± 5 per cent. A method is proposed which will enable most of the disadvantages of the older method to be overcome.