We have measured the stress birefringence in several magnetic garnets (YIG, YbIG, GdTbIG, and CVBIG) and one nonmagnetic garnet (GdGaG). Measurements were made at room temperature by applying both a static uniaxial stress and a saturating magnetic field along ⟨100⟩ and ⟨111⟩ axes. Nonuniformities in the samples and the loading procedure made it valuable to measure the birefringence automatically as a small light spot scanned the crystal. It was found that the photoelastic effect is 2 orders of magnitude too small to account for the linear magnetic birefringence as a birefringence associated with the macroscopic magnetostrictive strain. The stress optical constantB(&Dgr;n/stress) in the magnetic crystals was found to be about −1.8×10−13cm2/dyn for stress along ⟨100⟩, and −3.0×10−13for stress along ⟨111⟩. Diamagnetic GdGaG has values forBof −1.2×10−13and −3.0×10−13, respectively. It seems plausible then that the tensors describing the photoelastic and magneto‐optic effects are not strongly coupled. The dispersion of the stress birefringence is very slight in the range 4000–10 000 cm−1, implying that the photoelastic effect arises from the remote transitions responsible for the refractive index.