Methods for the measurement of the thickness of thin-foil specimens used in transmission electron microscopy are either difficult to carry out or have been subject to criticism. In particular, the contamination spot method is said to overestimate the thickness because the region of rapidly changing contrast marking the apparent edge of the spot is not on the foil surface but is on a broad contamination deposit whose thickness is changing much more slowly. A new method for measuring foil thickness is proposed, based on contamination deposits on the foil surfaces. The problems of the contamination spot method, in which the deposit is of circular form, are avoided by using one of the condenser lenses to focus the electron beam in a thin line on the foil during deposition. Adequate contrast can be obtained with a line whose width is one-third to one-fifth of the foil thickness and having a height equal to or less than its width. The error, being a fraction of the line width, is then very small. After rotation of the foil, the lines separate into two and the corresponding edges of the lines provide distinct features whose separation can be measured to determine thickness. The axis of rotation, perpendicular to which the separation of the lines has to be measured to calculate foil thickness, is determined by depositing two contamination lines at right angles. The method allows a number of measurements of thickness covering a relatively large area of foil to be made per contamination experiment. Near the edge of the foil, the upper and lower lines of contamination can join around the foil edge to form a U shape which can be used to measure thickness profile of the foil right up to the edge.