In an attempt to design a new stereo vision test we found that the spatial uncertainty effect had substantial impact on the measured stereo acuity thresholds. This effect is present whenever there is uncertainty of where the critical information occurs in a visual target. We studied the spatial uncertainty effect for stereo acuity and for line length estimation. In addition, we determined the temporal uncertainty effect in a stereoacuity test. In foveal vision, uncertainty effects increased visual thresholds, sometimes dramatically. For example, for an uncertainty factor of 25 (i.e., the critical information occurred randomly in any of 25 positions) stereo thresholds increased more than four-fold compared to those obtained without uncertainty. Although the uncertainty effect has been well described theoretically, we think that it is not always appreciated in the design of clinical or experimental visual tests. In tests aimed at determining visual thresholds, spatial and temporal uncertainty factors should be minimized.