AbstractLead–rubber bearings are seismic isolators that have been extensively used to protect buildings, bridges and mission‐critical infrastructure from the damaging effects of earthquake shaking. Testing of isolators, including lead–rubber bearings, is conducted either for research purposes (for example, shake table testing) or for determination of their properties before their use in applications (prototype and quality‐control testing). Testing is of significance as it (a) provides the data needed to validate complex models for analysis, and (b) results in basic mechanical properties of the tested isolators for use in analysis and design. Often testing has to be performed on reduced‐size bearings and/or at conditions of reduced speed of motion due, primarily, to unavailability of testing equipment capable of testing full‐size specimens at realistic dynamic conditions. When testing at reduced size, the conditions of testing need to be adjusted on the basis of scientific analysis so that the test results represent the behavior of the full‐size prototype bearing. This paper examines the issue of testing of lead–rubber bearings at reduced scale and presents the necessary conditions for testing in order to correctly reproduce the critical effects of lead core heating on the characteristic strength of the bearing. Experimental results on large‐size and reduced‐size bearings are used to demonstrate the concepts. Copyright © 2010 John Wiley & Sons, Ltd.