Franz Halder, chief of staff of the German Army High Command, had drafted the original 1939 plan for an invasion of France. It looked depressingly familiar - a heavy push through central Belgium toward the Somme, basically the Schlieffen reflex of 1914 with better trucks. Hitler disliked it. Halder's commander, Walther von Brauchitsch, was lukewarm. The senior staff officers who would have to execute it were openly pessimistic. Manstein, working under General Gerd von Rundstedt at Koblenz, kept hearing the same problem from his commander: the Halder plan could not actually deliver a decisive victory. Then, on 10 January 1940, a German aircraft carrying a partial copy of the plan crash-landed in Belgium - the Mechelen incident - and Belgian intelligence got an eyeful of the German intent. Halder rewrote the plan in places, but the basic concept did not change. Manstein, meanwhile, was already working on something else.

Manstein's instinct was to swing south of the main Allied line, push armor through the Ardennes - the hilly, wooded terrain along the Belgian-Luxembourg-German border that French doctrine had written off as impassable to mechanized forces - and break out toward the English Channel coast. His XIX Panzer Corps commander, Heinz Guderian, made the idea more radical. Don't even bother engaging the bulk of the Allied armies, Guderian argued. Race the panzers to the Channel, cut the supply lines, and the Allied forces in Belgium will collapse on their own. Manstein worried about the open southern flank such a dash would create. Guderian convinced him a secondary spoiling attack toward Reims could cover it. When Manstein presented the idea to OKH, Halder and Brauchitsch dismissed it as reckless. Manstein and Halder had been rivals for years - Manstein had been the obvious successor to chief of staff Ludwig Beck in 1938 before being pushed aside in the Blomberg-Fritsch affair. In late January 1940 Halder solved the immediate problem by promoting Manstein out of the way to command XXXVIII Corps in eastern Germany.

Manstein's staff in Koblenz did not take the dismissal quietly. Lieutenant-Colonel Gunther Blumentritt and Major Henning von Tresckow - the same Tresckow who would later become a leading figure in the July 1944 plot to kill Hitler - contacted Lieutenant-Colonel Rudolf Schmundt, Hitler's army adjutant and an old acquaintance of Tresckow's, when Schmundt visited Koblenz. Schmundt briefed Hitler on 2 February. Hitler had never liked Halder's plan and had been suggesting alternatives for months, some bearing a resemblance to Manstein's concept. On 13 February he ordered a change of strategy in line with Manstein's thinking, even though he had only heard a rough outline. On 17 February, Manstein was summoned to the Reich Chancellery to meet Hitler in person, with Alfred Jodl and Erwin Rommel present. Hitler took an immediate dislike to him - found him arrogant and aloof - but was impressed by the plan. After Manstein left, Hitler reportedly said: Certainly an exceptionally clever fellow, with great operational gifts, but I don't trust him. Halder issued the rewritten plan, Aufmarschanweisung Nr. 4, on 24 February.

When the attack came on 10 May 1940, Panzergruppe Kleist - five panzer divisions of Army Group A - drove through the Ardennes toward the Meuse River between Namur and Sedan, against second-rate French divisions in the Second and Ninth Armies. French strategic reserves, including the powerful Seventh Army, had been committed to the Breda variant of Plan D - a dash through the Low Countries to link up with the Dutch. The diversionary attack by Army Group B in northern Belgium worked exactly as intended. The French and British walked into the trap. On 13 May, German engineers and infantry forced crossings of the Meuse at Sedan, with Luftwaffe bombing collapsing the morale of the French 55th Division at the critical moment. Then Guderian and his fellow panzer generals did something the plan did not actually call for - they ignored their orders to consolidate and raced for the Channel. Hitler issued halt orders on 17, 22, and 24 May; the tank commanders mostly worked around them. Within ten days the panzers had reached the North Sea coast, splitting the Allied armies in two. Operation Dynamo evacuated 338,000 British, French, and Belgian soldiers from Dunkirk. The Belgian army surrendered. The remaining French and British forces lost the second campaign, Fall Rot, and Marshal Petain's government signed the armistice at Compiegne on 22 June 1940.

Historians have argued about this victory ever since. J.F.C. Fuller and Basil Liddell Hart claimed immediately after the war that the campaign proved the long-evolving German Blitzkrieg doctrine they had partly inspired. The Manstein plan, in this view, was a deliberate Revolution in Military Affairs. Later historians - Robert Doughty, Karl-Heinz Frieser, Adam Tooze - have pushed back. There is little evidence of explicit Blitzkrieg doctrine in pre-war German army records. German tank production had no special priority. The economy was geared for a long war. The Manstein plan, in this revisionist view, was a return to nineteenth-century Bewegungskrieg - war of maneuver - adapted to modern technology, less a doctrinal breakthrough than a calculated improvisation by men who knew Germany could not win a long war and bet everything on a single throw. Adam Tooze called the victory a risky improvisation. The concentration of armor in the Ardennes was an extraordinary gamble; if Allied bombers had hit the columns winding through those narrow roads, the entire operation could have collapsed into chaos. They didn't. The Wehrmacht got away with it once, and only once - the same Bewegungskrieg principle could not be repeated against the Soviet Union in 1941, because there was no equivalent decisive point to concentrate against. Manstein himself returned to eastern Germany after his February meeting with Hitler and took no further part in the planning. The campaign that made his name unfolded without him. More than 360,000 Allied soldiers were killed or captured in those six weeks. So did about 27,000 Germans. The plan worked exactly as designed, and the world it produced lasted just over five years before it began to come apart in the snow east of Moscow.