If the emission from ultraluminous X-ray sources (ULX) is anisotropic, the luminosities can be less than the Eddington Luminosity for a 15 Msun black hole (2E39 erg/s), weakening the need for intermediate mass black holes. We can test the isotropy condition because a fraction of the X-ray emission is absorbed by cold material, and much of this absorbed power will be reemitted isotropically by dust grains at infrared wavelengths. The ratio of the infrared luminosity to the absorbed X-ray luminosity is a measure of the anisotropy of the X-ray emission. We do not detect infrared emission from the ULX in NGC 1313 (X-2) using observations obtained with the IRAC and MIPS instruments on the Spitzer Observatory. For the ULX from Holmberg IX, IR emission is detected in two IRAC bands (3.6 μm, 4.5 μm) but not with MIPS. The detection and limits to the infrared power are well below the value expected if the X-ray emission is isotropic. Our model for the conversion of absorbed X-ray photons to dust reemission implies that the opening angle of the X-ray emission is less than 30 degrees (half-angle, two-sided jet), implying that the intrinsic X-ray luminosity is sub-Eddington.