The global transition from analogue to digital over-the-air TV broadcasting has freed up a large portion of spectrum in the United States and in other countries. Meanwhile, in response to the rapidly increasing traffic on wireless access networks the paradigm of shared spectrum, which allows secondary devices to opportunistically access licensed spectrum that is underutilised by primary users (PUs), is being actively explored. In TV bands, research has focused largely on enabling white space (WS) sharing, i.e. co-channel reuse outside the protection region. In this overlay mode, secondary users (SUs) must not operate in the protection region (which consists of the service region and its vicinity and is known as the no-talk zone) to protect primary users (e.g. TV receivers). The existence of no-talk zones is to protect PUs in the boundary service region. In contrast, grey space (GS) sharing is co-channel reuse inside the protection region. In this underlay mode, an SU needs to ensure that the additional interference it contributes is weak enough at nearby PUs to be deemed acceptable. This is facilitated by the presence of built environments within a protection region, i.e. if the secondary is indoors and the primary outdoors, we expect the availability of (indoor) GS and (outdoor, in the vicinity of the buildings) WS to increase, due to the attenuation effects that provide additional inherent protection to primary receivers. The goal of this article is to revisit interference protection criteria for outdoor WS, and extend them to outdoor GS and (further) indoor WS/GS. We carry out a case study for US TV bands to assess the (new) indoor–outdoor WS and GS availability. Our results show that there is more WS indoors than outdoors (as expected); we quantify this as a function of key parameters of the deployment scenario and (importantly) the attention effects of building walls. We also show that there are sufficient GS opportunities in outdoor environments, and many more indoors, leading to the conclusion that the often-repeated folk theorem that ‘white (and grey) spaces don’t exist where they are needed most’ needs substantial revision.