An interlaboratory study of the measurement of thermal diffusivity and thermal expansion characteristics of metal and polymer matrix composites has been conducted with the purpose of determining the effect of measurement technique on the result and the likely reliability of data.Reasonable agreement was obtained in most of the tests, although there remain some questions of reproducibility of thermal diffusivity results despite the use of "reference materials" with "recommended data" to determine apparatus performance. PMCs proved difficult to test in representative thicknesses because of their low thermal diffusivity. Reducing the thickness to a level where the rear-face thermal transient is more readily measured would result in testing only one or two plies which may not be representative of the full normal layup thickness.In thermal expansion measurement, composite materials are shown to exhibit progressive size changes with 1hennal cycling. In part this is due to annealing effects on the first cycle as the test-piece adjusts to the heating/cooling rate of the tests, and to ratchetting effects due to plastic defonnation in the matrix phase. When thennal expansion tests are made with a push-rod dilatometer, the added push-rod force appears to result in additional contraction effects, especially during the first thermal cycle, compared with inteferometric tests in which the applied stress is much lower.