Accuracy of an open source remote eye tracking system and a state-of-the-art commercial eye tracker was measured 4 times during a usability test. Results from 9 participants showed both devices to be fairly stable over time, but the commercial tracker was more accurate with a mean error of 31 pixels against 59 pixels using the low cost system. This suggests that low cost eye tracking can become a viable alternative, when usability studies need not to distinguish between, for instance, particular words or menu items that participants are looking at, but only between larger areas-of-interest they pay attention to.