it’s 6:40pm. the pajamas are still on the arm of the couch where you put them twenty minutes ago.
your eleven month old is running the same stretch of hallway, back and forth, laughing at nothing. he pulls up on the coffee table, drops down, gets back up. when you crouch to catch him he arches away from you, delighted, and takes off again. his cheeks are pink. his eyes are bright and a little glassy and he will not sit still for anything, least of all you.
most of his dinner is on the floor. the laugh coming out of him is slightly too loud for the size of the room.
nothing about this looks like tired.
so you make the reasonable call, the one most people make, the one i would have made before i knew better: he’s clearly not tired, we’ll try again at 8pm.
and at 8pm he fights you for forty-five minutes, goes down hard, and is up at 4:50am like none of it happened.
the short version, for the tired and the skimming: when i’m not sure, bedtime goes earlier. not later. overtired costs more than undertired, and those two mistakes are not the same size. this is the single change i recommend more than any other, and it still will not fix everything, which is the second half of this letter.
the part i already wrote down
back in june i wrote the mechanism out properly, in why your baby fights sleep harder when they’re overtired. i’d rather point you there than run it again, so here’s the one paragraph you need to follow the rest of this.
a child who stays awake past the point their body wanted to sleep starts producing cortisol. that’s the stress hormone, and it’s also the one that’s supposed to peak in the morning and get you out of bed. so the most tired child in the room is often the most wired one. more tired looks like more awake. it isn’t a figure of speech, it’s a hormone doing exactly its job at the wrong hour.
which means the hallway laps are not evidence he has hours left in him. they’re the opposite, and they’re the reason the 8pm bedtime went the way it did.
if that’s new to you, read that letter first. this one is what i actually do about it.
the two mistakes are not the same size
this is the real reason for the bias, and it’s the part almost nobody says out loud.
put him down twenty minutes before he was ready, and here is what it actually costs you. he lies there. he rolls onto his side, pulls himself up on the rail, talks to the wall, kicks the mattress twice for the sound it makes. sometimes it’s twenty minutes. sometimes it’s longer and genuinely maddening, and there’s a version of too-early that turns into its own fight, which i’ll come back to at the end. but whatever it costs, it’s over when he falls asleep. tomorrow starts clean.
put the same baby down at 8pm, well past his window, and it doesn’t end there. it takes him longer to fall asleep, not shorter. the surfacings between sleep cycles, which every child has and always will, are more likely to turn into a full wake-up, because sinking back down is harder with cortisol on board. then the 4:50am wake starts the next day short. a short day pushes the next bedtime late. that’s how one rough tuesday turns into three rough weeks, and it’s why the version of him you met at 6:40pm keeps showing up.
one mistake ends when he falls asleep. the other one follows you into the morning, and then into next week. so when i genuinely can’t tell which side of the line a child is sitting on, i take the one that ends. that’s the whole bias, and that’s all it is.
what it looks like from the floor
these are the things i’m listening for when a parent describes their evening. you have already met most of them.
more energy at 6:30pm than at 5:30pm, and if you watch it closely it’s frantic rather than happy. the hallway laps. the laugh that’s pitched a little wrong.
arching away from you when you reach for him, and finding it hilarious
bedtime taking an hour, with real protest in it
the drop and jerk: they start to go under, jerk awake, fuss, resettle, repeat
a wake before 5am that looks like a genuine morning
not one of those reads as “tired” to a parent standing in the room at the time. that’s the whole trouble with it. the first four are on my overtired list without much hesitation. the last one is on it too, but it’s also on another list, and i’ll come back to that.
how i actually move it
here’s the part that gets skipped, because “put them down earlier” is said constantly and explained almost never.
i don’t really move bedtime. i move the wake window and let bedtime land where it lands.
the principle i build every plan on is a rhythm rather than a schedule. bedtime is the end of the last wake window of the day, whatever time that lands on.
so rewind that evening. the afternoon nap fell apart at thirty minutes, which nobody thinks about again at 6pm, but it means the last stretch of his day started earlier than usual and has been running longer than it looks. the day was shorter. the window ran out sooner. bedtime moves earlier, and not because you decided to get strict about it. you’re following his actual day instead of overriding it.
in practice that means the question at 6:15pm isn’t “is this too early?” it’s “when did he last sleep, and how long has he been up?” wake windows are a range, not a number, and where you land inside that range depends on how the last nap really went. a chart can’t see your child’s afternoon. you can. you were there.
then keep the sequence identical no matter what hour it starts at. feed, wind-down, down. lights low and warm rather than white or overhead. the routine is the cue, not the clock, and that’s what lets a 6:15pm bedtime feel like bedtime instead of feeling like something has gone wrong.
now the limits, and there are three
one. it will not fix everything. if you bias earlier for four or five nights and nothing moves, stop and look elsewhere. the clean cause is something else, and running the same change harder is how families lose weeks they didn’t have. an earlier bedtime does nothing for a child who can only fall asleep being fed, for instance. that’s a different letter.
two. earlier is where i go when i’m unsure. it isn’t somewhere to keep going. here’s the honest version, including the part where my own training pushes back on the title of this letter. the curriculum i trained in has a rule that applies to every method in it, and it’s called “don’t automatically move bedtime earlier.” i agree with it. the word carrying the weight is automatically, and the rule’s own next line is “unless baby is truly ready to sleep.”
that’s the whole distinction, and it’s worth being precise about. sliding bedtime earlier by the clock, as a reflex, because earlier is generally better, is the thing the rule warns against, and it warns against it for a real reason: the drive to sleep builds across the day, and a bedtime that lands before your child has built enough of it hands you the same fight from the other direction. a small person lying in the dark, wide awake and cheerful, for forty minutes, and then a bedtime that gets harder rather than easier, because the room has become somewhere you lie awake. that’s the version of too-early i said i’d come back to.
what this letter is describing is the other thing: a read of one specific child on one specific evening who is showing you the tells above. that’s the “truly ready” half of the rule, not an exception to it. earlier is a tiebreaker when a child is telling you he’s ready, not a standing instruction.
and the rest of that rule is about timing, which is the part i’d hold hardest. keep bedtime stable while you’re actually running a plan. if you’re partway through one, this is not the week to slide bedtime around. change two things at once and you learn nothing from either one. bias earlier before you start, or after the plan has settled, not in the middle of it.
three. the early morning wake is the one that fools everyone, because two opposite things cause it. a bedtime landing too late, or a body clock that has genuinely decided the day starts at 5am. an earlier bedtime helps the first and does nothing at all for the second. which one you’ve got changes the answer completely, and you usually can’t tell from a single night.
and the line underneath all of it: if the timing changes don’t touch anything, and there’s pain or feeding or breathing anywhere in the picture, that’s your pediatrician and not me. reflux, weight, congestion, feeding volumes, always theirs. i’m clear about where my lane ends, and that’s most of why you can trust what’s inside it.
the question worth swapping in
if you take one thing out of this letter, take the question you ask yourself at bedtime.
not “why is he fighting me on this?” but “when did he last sleep?”
the first one has no answer in it and a lot of blame. the second one has an answer, and it’s almost always sitting on the clock behind you rather than in anything you did.
so, tonight, when he’s doing his laps: what time did he actually last sleep, and what would bedtime look like if you counted forward from there instead of from the clock?
if this is the kind of thing you want in your inbox, the letters are free and they land whenever i write one.
if you’re past the point where one change is going to do it, the way i work with families starts with a 10-minute intake. it’s how i see your whole picture before we ever talk. tell me about your child. no pressure, the letters aren’t going anywhere.
✶ and for the record, you’re doing great.



