This paper describes analytical and experimental thermal results for a 33.5 cm diameter plano mirror under conditions of small thermal perturbations (steady-state temperature gradients of 10–100 mK). These tests are intended to support verification of specific thermal requirements for a space interferometer. The primary thermal requirement is knowledge/control of temporal changes in mirror gradients to the 1 mK (0.001 K) level. Tests were performed with small heat inputs to the back of the mirror, which was suspended in a thermally-uniform shroud. Correlation of thermal models for both conductively and radiatively-heated test configurations were performed, and results indicate very good agreement between the thermal model predictions and the temperature measurements. The modeling uncertainty based on the test correlation is estimated to be ±3 mK for absolute temporal comparisons. Comparisons for temporal change of gradient are shown to be within 1 mK for small perturbations. The paper describes the test setup, test results, model correlation and uncertainty estimates.