The final chapter brings together the various strands discussed so far. As acknowledged throughout the book, SRM is ethically problematic in some deeply disturbing ways. But given that we should consider climate justice through the lens of non-ideal theory (see chapters two and three), and given that we should not view cases in which SRM is potentially attractive as genuine moral dilemmas (see chapter five), there may be situations in which deployment of some SRM policy is ethically permissible. This is plausible if we take a comparative approach to ethical assessment of climate policies. An ethically problematic climate policy can be permissible if it compares favorably to other available options. In a case in which no available policy avoids having serious ethical problems (a pessimistic scenario), it is not enough simply to point out some policy’s problems and conclude that it is therefore prohibited, for this would require us to deem all the available policies to be prohibited, yielding a genuine moral dilemma. The question of what we ought to do in a pessimistic scenario is a pressing one, calling for moral action-guidance—something that a moral dilemma framing would foreclose. Unfortunately, a pessimistic scenario is precisely what we can expect without sufficient reductions in greenhouse gas emissions. At some point, past emissions could commit us to very dangerous climate change. In such a scenario, all courses of action carry ethical problems: emissions mitigation cannot address committed climate change; adaptation has limits and may be very expensive; and business-as-usual would simply permit the harm and injustice of climate change to accumulate. Despite its own ethical problems, there are reasons to think that some policy involving SRM would compare favorably to other options in certain pessimistic scenarios. Because SRM has the potential to cool the climate quickly, it can manage some climate risks better than slower-acting responses, such as mitigation and adaptation, a major advantage when it comes to securing beneficial outcomes while minimizing harmful ones (see chapter one). Moreover, while SRM is expected to yield distributive and procedural injustices (see chapters two and three), I argue that these are tolerable from a non-ideal-theoretic perspective if they are less severe than the injustices of all other competing policies, and I present reasons for thinking this may be the case with SRM. Importantly, I do not defend the view that an “SRM-only” policy would be attractive in pessimistic scenarios. I sketch a hybrid policy involving short-term SRM to provide time for emissions mitigation and adaptation. This has advantages over an SRM-only policy. To take one example, it involves less intergenerational injustice, because it does not put future generations indefinitely at risk of abrupt termination of SRM, which would result in rapid, harmful warming.