The one-paragraph verdict

Andrew had something almost no historian gets: the actual files. The result is the most credible single account of MI5 you can buy, and the writing is far better than its bulk suggests — engaging where it could easily have been a chronology. But it is a thousand pages, and it reads like it. The acronyms come thick and there is no glossary, so you either already know the cast or you keep a tab open. A couple hundred pages lighter and it would be a recommendation instead of a maybe.

Who should read it

Read it if you care about 20th-century British history, the mechanics of counter-espionage, or how an institution that exists to keep secrets nonetheless generates a paper trail. Skip it if you want a brisk spy narrative — this is institutional history, not a thriller, and the official sponsorship means it is candid but never gossipy.

Where it earned its place

No platform tie-in here; this is pure off-duty reading. What it left me with is a healthier respect for primary sources over the secondhand version of events — the gap between what gets reported and what the records actually say is the whole point of the book, and a useful habit of mind to carry back to work.

Skip it if…

If acronym-dense, source-heavy institutional history sounds like a chore rather than a pleasure, this will confirm your fears. Length is a real cost here, not a quibble.

[The Defence of the Realm on Goodreads]