Line of business restrictions (LOBRs) are antitrust remedies or regulatory restrictions that limit the activities that a firm can undertake. They include separation restrictions ranging from structural to behavioural separation (accounting, functional or legal). However, there are also alternative behavioural restrictions such as mandating access, non-discrimination obligations and mandatory standards on portability and interoperability. LOBRs are used to address concerns that competition is likely to be prevented, restricted or distorted in a number of different ways. These include concerns that dominance is leveraged to exclude rivals in another downstream market, for example through discriminatory selfpreferencing (equivalent to margin squeeze) or refusal-to-deal. The risk of such exclusionary abuses also arises from mergers that might make such conduct likely. Each type of LOBR has enjoyed success in certain circumstances; the challenge is therefore identifying the right LOBR for the specific problem. For example, if the theory of harm is that access is being denied to an essential facility, and a loss of competition as a result of such an explicit refusal-to-deal is established, then the solution will need to focus on allowing access, and applying a duty-to-deal. Hence, the appropriate LOBR might include mandatory access on FRAND terms, or structural separation to achieve the same access. In contrast, if the regulatory or antitrust concern is not that the firm refuses-to-deal, but that it deals on terms that foreclose through raising rival costs or predation (via some form of margin squeeze, or equivalently a merger leading to margin squeeze), then the solution (were one required) would not focus on mandating access. Instead, the response might be a non-discrimination obligation (without necessarily imposing a duty-to-deal on FRAND terms) or to mandate standards for portability and interoperability. Notably, in the case of digital platforms, it would appear that refusal-to-deal is not driving the concerns (thus far), and so imposing a duty-to-deal seems unlikely to be the right answer. Instead, the concerns appear to relate to foreclosure through self-preferencing that raises rivals’ costs, meaning that non-discrimination obligations are the more likely LOBR. Moreover, digital platforms are not natural monopolies in which competition cannot take place. Therefore, the approach should not be to write off the possibility of competition in platform markets and focus solely on preserving the possibility of competition downstream (or upstream) markets (as in regulated infrastructure monopolies). Instead, behavioural LOBRs including standards for portability and interoperability could facilitate competition within these markets (and thereby begin to resolve the foreclosure risk in downstream markets). However, it may nevertheless be sensible to reinforce these with further LOBRs, for example non-discrimination obligations, in these downstream (upstream) markets to preserve competition while competition in these ‘core’ markets is strengthened.