It all depends when the breach occurs.  So, Adam states "The choice whether to affirm or terminate occurs after (upon) the breach and the duty to mitigate has arisen." I don't think it does because without the acceptance of the repudiation there is no breach at all. It is writ in water. That is why a repudiation can be withdrawn. If it were itself a breach without more it could not be.  As to the cases Asamera v Sea Oil did not concern a repudiation, but an actual breach Habton v Nimmo concerned a breach of warranty of authority, not repudiation. And so there was a duty to mitigate from the moment of the wrong. Dunno about the Saskatchewan one Steve asks  "If you fail to deliver goods to me and, without having in any way indicated that I am no longer expecting you to perform, I later sue you, I will not be given specific performance (unless the goods are unique) and my damages will be calculated assessed on the basis of the market price as of the contractual date of delivery. Yet if I did not terminate, then according to the orthodox understanding of the relationship between termination and mitigation (as set out by Rob and Andrew), should not damages be calculated as of the date of judgment?" The plaintiff is entitled to damages reflecting his right to performance, assessed at the time of breach (here time of non-delivery). Consequential loss is irrelevant to this head of damages. If he wishes to argue that he has loss consequent upon the breach of the duty to deliver on a particular day over and above the difference in value this is recoverable (although usually difficult to show where there is a ready market for the goods that he can avail himself of, ie he either can or should mitigate). Consequential losses are assessed at time of trial. Termination does not come into it.  I tried to explain the law on timing of damages in sale cases (amongst other things) in a book chapter edited by one J Neyers. Available to be read here I see http://denning.law.ox.ac.uk/news/events_files/A_Golden_Victory_or_Not.pdf