The 'essentially contested concept' of security is analysed, and some main kinds of ambiguity and dimensions outlined: level, kind of threat and kind of defence. Discourses on security, particularly of small nations, must avoid being trapped into dealing only with one level (national, which in practice normally means state), one kind of threat (military) and one kind of defence (again military). There is no clear relation between kind of alignment and military expenditures, but non-aligned states are overrepresented both among the very high armers and among the very low armers. Increasing gaps in military R&D make it necessary for small states to rely more on different military strategies from the great powers, putting more stress on invulnerability and on non-offensive defence. For bi- or multi national states, the relation between nations and state becomes crucial for any defence, and in particular for social and cultural defence, which presupposes the absence of major grievances that can be exploited by external powers. In welfare terms, but also in terms of national independence, economic threats are often at least as serious as military ones One remedy is diversification of export commodities and of trade partners. Collective strategies of small states suffer from the 'free rider' problem, and from the vulnerability to collective counter-strategies from centre actors. It is therefore wise of the Non-Aligned Movement not to have attempted to become a military actor. Collective economic strategies will have to take other forms than the producers' cartel, whose difficulties are almost always forbidding.