Many previous papers have described well-developed simulation programmes for electromagnetic launchers, many of which use a coupled-circuit filamentary approach in describing the electrical circuits during the launch period. However, the accurate validation of these programmes presents several very significant problems, due to certain mainly mechanical effects that are extremely difficult to quantify in any relatively straightforward and accurate manner. The present paper explains these difficulties, and describes an experimental investigation in which their influence on the results is substantially absent, enabling close agreement with results predicted from a filamentary model to be achieved.