Virtual Reality(VR) systems have the potential to revolutionise how we interact with computers. However motion sickness and discomfort are currently severely impeding the adoption. Traditionally the focus of optimising VR systems have been on frame-rate. Delay and frame-rate are however not equivalent. Latency may occur in several steps in image processing, and a frame-rate measure only picks up some of them. We have made an experimental setup to physically measure the actual delay from the user moves the head until the screen of the VR device is updated. Our results show that while dedicated VR-equipment had very low delay, smartphones are in general not ready for VR-applications.