So here is the direct answer before any of the detail. Lock as soon as you are under contract on a specific house, because that is the first moment you have a real closing date worth protecting. Getting under contract in the first place usually takes a verified preapproval rather than a prequalification. And lock for a longer window than you think you need, because a few extra days of coverage usually costs a little, while running out of lock can cost a lot. Everything below is why that answer holds, and what it costs you when you get it wrong.
A lock is a promise with a date attached
A rate lock is a promise from the lender. When you lock, the lender agrees to hold a specific interest rate for you for a set number of days while your loan is processed, no matter what the market does in between. On an adjustable loan, what you are locking is the starting rate, so how long that starting rate lasts and what governs it afterward is a separate question from the lock.
Mortgage rates move constantly, tied to markets that shift daily and sometimes shift twice in the same day. Without a lock, the rate you were quoted is nothing more than a snapshot of the moment you asked. The lock freezes it, so the market cannot move it out from under you in the weeks between your offer and your closing table.
It matters because the stretch between contract and closing is exactly when you are least able to absorb a surprise. You have committed to a house. You have movers to schedule and a lease to end. That is the worst possible time to find out your payment went up. The lock removes that uncertainty from the part of the process where uncertainty hurts most.
How long does a rate lock last
Locks come in windows. Thirty, forty five, and sixty days are the common ones, and longer windows generally cost a little more, because the lender is carrying the risk of the market moving for a longer stretch on your behalf.
The right window is not the cheapest one. It is the one that comfortably covers how long your loan will take to close, with cushion on top. If your contract says thirty days to close and you lock for thirty days, you have given yourself exactly zero room. Zero. A slow appraisal eats the whole cushion, and so does a title problem. An appraisal is the independent estimate of the property's value that the lender orders from a licensed appraiser, and it can sit in a queue for a week you did not budget for. A title problem is something unresolved in the legal record of who owns the property and what claims are attached to it, and those get discovered late by their nature. Add a seller who needs an extra week, and a short lock runs out before you ever reach the table. Where the time actually goes between an accepted offer and closing day is the map to build your lock window against.
Ask your loan officer a plain question: given this file, this property, and this closing date, what lock length gives us real breathing room. A good answer names a realistic timeline, not an optimistic one.
What happens if rates move after you lock
This is the part people get backwards, so here it is in the simplest form.
If rates rise after you lock, you keep the rate you locked and the increase does not reach your loan. That is precisely what the lock was for, even if no money changed hands to get it.
If rates fall after you lock, you are generally held to what you locked. The promise runs both directions. The lender committed to holding your rate, and you committed to taking it.
There is one common exception, and it is worth knowing about before you assume you have it. Some lenders offer a float down, a feature that lets you capture a lower rate one time if the market improves meaningfully after you lock. It usually costs something, and it usually comes with rules about how far the market has to move and how late in the process you can use it. Find out whether your lock has a float down attached, what it would cost, and what conditions trigger it, before you find yourself watching the market and hoping.
The expiration date, and the fee nobody plans for
Every lock has an expiration date. It is on your paperwork. Most buyers never look at it.
If your closing drags past that date, the lender may require you to pay to extend the lock, or the rate may be repriced to whatever the market is doing that day. Extension pricing varies by lender and is often quoted as a small fraction of one percent of the loan amount, which on a loan the size of a mortgage turns into real money for nothing new. It is priced the way points and the rest of the lender's own charges are priced, as a slice of the loan amount. Same rate, same house, same loan. You paid for time. So ask, in writing and early, what an extension would cost in dollars on a loan your size. That answer is easy to get in week one and painful to discover in week seven.
Here is the part that should get your attention. Blown locks are usually not caused by the market. They are caused by a slow file. An appraisal ordered late. A bank statement the borrower took nine days to send. A letter of explanation nobody chased, meaning a short signed note from you giving the lender the story behind something in your file, such as a deposit or a gap in employment. Those requests come out of underwriting, and knowing what conditions look like before they land is how you clear them in a day instead of a week. That is why the boring advice, answer every document request the same day you get it, is not your loan officer being pushy. It is the cheapest protection your locked rate has.
Locking versus floating, and how to decide
Floating means deliberately not locking yet, betting that rates will improve before you close. Locking means choosing certainty over a bet.
I do not predict rates, and you should be careful with anyone who does. The honest framing has nothing to do with where the market is headed and everything to do with two questions about you. First, how long is your timeline, and how much room does it have in it. Second, if rates moved against you by a meaningful amount before closing, would that be an inconvenience or would it break the deal.
If a move against you would break the deal, you are not in a position to gamble, and the certainty is worth more to you than the possible upside. If you have a long timeline, real financial cushion, and you genuinely understand that you are placing a bet, floating is a legitimate choice made with open eyes. What is not legitimate is floating because it feels like waiting for something free. It is not free. It is a position, and positions can go against you.
If you want to know what that position is actually worth, spend five minutes sizing it with your own file rather than reading a number of mine. Rate and payment figures move constantly, so I am not going to print any here. Take the loan amount you are looking at and a rate you have been quoted by a real lender, put both into the mortgage payment calculator, and write down the monthly figure. Then run it again with half of one percent added and nothing else changed. Two things usually surprise people. The monthly difference is smaller than they expected, and the total interest difference across the full term is very much larger than they expected. That gap between the small monthly number and the large lifetime number is the entire reason a lock exists, and seeing it in your own arithmetic teaches it better than any figure I could put on this page.
A 45 day lock, a closing on day 50, and a bill for the difference
Say a buyer locks for 45 days and expects to close on day 35. Two things can happen.
In the first branch, the market moves against them a week before closing. Because they locked, they close at the rate they locked. They keep money they never see leave their pocket, which is why this kind of win feels like nothing at all. It is still a win.
In the second branch, the appraisal comes back late, the file stalls, and closing slips to day 50. That is five days past a 45 day lock. Now the buyer either pays to extend or takes whatever the market says that week. Same lock, same buyer, same house. The only difference was how fast the file moved.
The lesson sits right there. Match the lock to a realistic timeline, then keep the process moving so you never need the extension in the first place.
The myth: my quoted rate is locked in automatically
It is not. A quote is a number the market gave you on a given day. Until you formally lock, that number can move, and no amount of having it in writing in an email changes that.
The second half of the myth is just as common. People who do lock often assume the lock is permanent. It is not. It has an expiration date, and past that date you are back in the market. Knowing the difference between a quote, an active lock, and an expired lock is most of what keeps your rate from quietly slipping away.
What to do next
If you are already under contract, send one message today and ask for the exact expiration date of your lock in writing. Put that date in your calendar. Then put a second reminder five days ahead of it, because five days is enough time to chase a missing document and not enough time to be surprised. Most buyers find out the expiration date exists on the day they blow past it.
If you are not that far along, save three questions in a note on your phone and ask them the first time a lender talks to you about locking.
- Given this file and this property, what lock length covers a realistic closing timeline with room to spare.
- Is a float down available on this lock, what would it cost, and what has to happen in the market before I could use it.
- If we run past the expiration, what does an extension cost in dollars on a loan my size.
Three answers to three questions, and you will understand your own loan better than most buyers do at the closing table.
For the loan officers reading this
If you are the buyer, save the three questions and stop here. The rest is for originators.
Set the lock to a realistic timeline, not an optimistic one. The extension fee from a stalled file is the cost clients remember and resent, and they attach the memory to you, not to the appraiser.
Explain float down in real terms before it becomes relevant. What it costs, what has to happen for it to trigger, and what it could realistically save. A client who understands the feature in week one does not feel cheated in week five when the market dips and they cannot chase it.
Drive document turnaround relentlessly, because most blown locks are operational, not market driven. And never let a client float without making sure they know they are placing a bet, not waiting for something free. Saying that out loud is uncomfortable for about ten seconds and saves the relationship if the bet goes wrong.