The arm that has gone numb
It is the middle of the afternoon and you have not moved in forty minutes. Your baby is asleep on your chest, breathing in that slow, milk-drunk rhythm, and you are calculating—the way parents learn to calculate—exactly how many degrees you can shift your weight before the spell breaks. You lower them toward the crib. They are limp, deeply gone. The instant their back touches the mattress, their eyes fly open and their arms fling outward as if you had dropped them from a height. And you are back where you started, holding a baby who, it seems, will only sleep on you.
This is one of the most quietly exhausting experiences of early parenthood, and almost everyone goes through some version of it. It also comes wrapped in a layer of worry: that you are making a rod for your own back, building a habit you will pay for later, that a baby who only sleeps when held is a problem to be corrected.
Mostly, it isn't. What you are watching is biology doing exactly what it evolved to do. Understanding the mechanism doesn't magically make the numb arm comfortable—but it changes how you read the situation, and it points toward the small adjustments that actually help.
A baby's nervous system is built to be regulated from the outside
Newborns arrive with a nervous system that cannot yet steady itself. An adult who feels stressed can, with effort, slow their own breathing, talk themselves down, find calm. A newborn has almost none of that machinery online. Their capacity to move from alert distress back to a settled state depends, in the first months, on borrowing yours. Researchers call this co-regulation: the caregiver's calm body literally helps organize the baby's.
The channels for this are concrete. Skin-to-skin contact and being held trigger the parasympathetic nervous system—the branch responsible for "rest and digest"—which slows heart rate and lowers the stress hormone cortisol. The warmth of your body helps with thermoregulation, since newborns are not yet efficient at holding their own temperature. The gentle pressure of being contained against a chest mimics the close boundaries of the womb. Even your scent and the muffled sound of your heartbeat are signals the baby has known since before birth.
From the baby's point of view, then, sleep on your body and sleep in an empty crib are not the same task with two locations. They are physiologically different states. On you, the nervous system is being held steady. In the crib, it has to do that work alone—work it is genuinely not yet equipped to do well.
The fourth trimester, and why the crib feels wrong
The pediatrician Harvey Karp popularized a useful frame: the fourth trimester. Human babies are born, in a sense, three months early compared with other mammals, because a brain large enough to fit through the birth canal has to finish much of its growing on the outside. The result is a creature who is still, developmentally, expecting the conditions of the womb—constant motion, snug containment, sound, warmth, the steady presence of another body.
A flat, still, quiet crib offers the opposite of every one of those cues. It is not that your baby dislikes the crib. It is that the crib does not yet read as a place where sleep is safe, because every signal their ancient wiring is scanning for—am I held? am I warm? is someone near?—comes back negative. For most of human history, a sleeping infant set down alone and motionless was a vulnerable one, and the nervous system has not gotten the memo that times have changed.
This is also why the problem tends to ease on its own over the first several months. As the brain matures, babies grow better at self-regulating, more tolerant of stillness, more able to bridge the gap between your arms and the mattress. The held-only phase is, for most families, a phase.
The startle that undoes everything
Then there is the specific, maddening moment of the put-down: the limp baby who jolts awake the second they hit the sheet. That is usually the Moro reflex, sometimes called the startle reflex. It is an involuntary, hardwired response present from birth, in which a sensation of falling or a sudden change in head position makes the baby fling their arms wide, then pull them back in, often with a cry.
The Moro reflex exists for a reason—evolutionarily it is thought to be a grasping response, a way for an infant to catch hold of a caregiver during a fall. But it means that the very motion of lowering your baby and removing your supporting hands registers, to their brainstem, as falling. The reflex fires. The baby wakes. It is not stubbornness and it is not a verdict on your technique. It is a spinal-level reflex doing its job at the worst possible moment.
The Moro reflex typically fades on its own somewhere around three to six months as the nervous system matures. Until then, the trick is to keep it from triggering during the transfer.
Working with the biology instead of against it
None of this means you are doomed to a numb arm forever, or that you must simply wait it out. Once you see the put-down as a problem of preserved cues and an untriggered reflex, the workarounds become obvious.
Wait for deeper sleep. Babies cycle through light and deep sleep, and the first stretch after they drift off is light and easily disturbed. Give it ten to twenty minutes. Limp arms, an open hand, slow breathing—these signal the deeper sleep where a transfer is far more likely to survive.
Keep them contained. A snug swaddle (used safely, and stopped as soon as your baby shows signs of rolling) dampens the Moro reflex by keeping the arms from flinging wide. It also preserves the womb-like sense of being held after your arms are gone.
Go down feet first, slowly, and stay. Lower the lower body first, keep your head close, and leave a firm hand resting on their chest or belly for a moment after they are down. The pressure tells the nervous system that contact has not actually ended. Lift your hand only gradually.
Mind the temperature gap. Part of what wakes a baby is the shock of a cool sheet against a body that was pressed to your warm one. Warming the mattress slightly beforehand—with a hand, never with anything left in the crib—softens the contrast.
And—worth saying plainly—sometimes the right move is to let the contact nap happen. A held nap is real sleep. If you are at the end of your reserves and the transfer keeps failing, a baby asleep on a supervised, awake adult is not a wasted nap or a ruined habit. It is rest, for both of you.
Timing is the quiet variable
There is one factor underneath all of this that gets overlooked. A baby put down at the right moment—genuinely sleepy, but not yet tipped into the overtired state where cortisol surges and everything becomes a fight—is dramatically easier to transfer than one who has been kept up too long. Overtiredness makes the nervous system jumpier, the Moro reflex more trigger-happy, the whole settling process more fragile. So much of the held-only struggle is really a struggle of catching the window before it closes.
That window is maddeningly hard to read by guesswork, because it shifts as babies grow and varies from one to the next. This is the problem Drowsy was built for: it learns your particular baby's pattern and tells you the specific next window to put them down—so you are attempting the transfer when their biology is on your side, not against it. It won't switch off the startle reflex or rush the fourth trimester. But it takes the hardest variable, the timing, off your shoulders, so the numb-arm afternoons grow rarer on their own.
If you'd like a calmer way to find that window, you can meet your baby's rhythm at drowsy.lumenlabs.works.