Here is what most parents conclude after a failed sleep training attempt: their baby is one of those babies it does not work for. Their baby is too sensitive, too stubborn, too attached, too young, too old, too something. Sleep training works for other people’s babies but not theirs.
This conclusion is almost always wrong.
Sleep training works. The research on this is not ambiguous — multiple randomized controlled trials have followed sleep-trained babies into childhood and found improved sleep for babies, improved sleep for parents, improved maternal mental health, and no measurable negative effects on the children. The method is not the variable that determines success or failure. The execution is.
What fails is not sleep training. What fails is a specific set of mistakes that are so common, so predictable, and so understandable given the context of exhausted decision-making, that most families make at least one of them. Some make all five.
This article names every one of them — what they are, why they happen, what they cost you, and exactly what to do instead. If you have tried sleep training and it did not work, something on this list is almost certainly why. If you are about to start, this article is the preparation that most sleep training guides skip entirely.
Why Sleep Training Fails More Often Than It Should
Before the mistakes themselves, one piece of context that reframes everything.
Dr. Craig Canapari, MD, director of the Yale Pediatric Sleep Center and one of the most widely cited pediatric sleep specialists in the country, has written that if he had to narrow it down to the two most common reasons why parents struggle and fail with sleep training, inconsistency and the extinction burst are the most significant.
Those two forces underlie every single mistake on this list. Every mistake either creates inconsistency, forces a parent into the extinction burst unprepared, or both. Understanding that going in means you can see each mistake for what it actually is — not a character flaw, not a sign that you are doing something wrong as a parent, but a predictable pattern that has a predictable fix.
Mistake 1: Starting Before Your Baby Is Ready
This is the mistake that poisons everything that comes after it.
From zero to four months, babies lack the ability to self-soothe, and it is important for them to get the help needed to soothe to sleep. When a baby younger than four months is forced to sleep train, not only can the attempt fail to provide better sleep, but the baby and parents may experience extreme amounts of stress.
The damage here is not just that the attempt fails. It is that the failure becomes evidence. Parents try sleep training at six weeks because they are desperate and someone told them it worked for their baby. It fails — of course it fails, the neurological hardware for independent sleep onset does not exist yet. The parent concludes that sleep training does not work for their baby. They carry that conclusion into the four-month attempt, and they approach it with less commitment, less consistency, and more readiness to abandon it. The early failure creates the later failure.
Four months is the absolute minimum age, and even then, readiness is not guaranteed by a date on the calendar. It depends on weight — most pediatricians recommend at least 14 pounds — and on medical clearance. Before beginning any sleep training, a pediatrician should rule out reflux, ear infections, allergies, and sleep apnea. Each of these conditions disrupts sleep in ways that no behavioural method can override, and trying to sleep train a baby with undiagnosed reflux is not just ineffective — it is genuinely unfair to the baby, who is waking because they are in pain, not because they have not learned to self-settle.
Signs that your baby is genuinely ready beyond age and weight: they can stay awake for age-appropriate wake windows without immediately collapsing into overtiredness, they show consistent tired cues you can recognize and act on, they have at least one nap per day that lasts longer than 30 minutes, and they have been cleared by your pediatrician with no outstanding medical concerns.
If you start too early and it fails, do not conclude that sleep training does not work for your baby. Conclude that you started too early. Wait until your baby is ready and start again with a full plan.
The fix: Get your pediatrician’s explicit sign-off before your start date. Not just a general “yes you can try sleep training” — a specific conversation about your baby’s weight, feeding schedule, any symptoms that might indicate a medical cause, and whether they still need night feeds. This conversation takes ten minutes and removes an entire category of failure before you start.
Mistake 2: Inconsistency — The Single Biggest Failure Point
This is the mistake Dr. Canapari names first because it is the most common and the most destructive.
Inconsistency can stop any progress being made. Your baby is looking to learn and needs to understand what to expect at night.
The mechanism behind why inconsistency is so damaging is important to understand. The most powerful form of conditioning is inconsistent positive reinforcement — like pulling the lever in a slot machine. Inconsistency can actually favour the crying behaviour, as the child receives intermittent reinforcement. What this means in practice: a baby who is left to settle on Monday, rocked to sleep on Wednesday because it was a hard night, and left again on Friday is not confused — they are learning something very specific. They are learning that if they cry hard enough and long enough, the response eventually changes. Their brain is optimising for the behaviour that sometimes produces a reward. That is not a parenting failure. That is how conditioning works, and it works exactly the same way in every human regardless of age.
The most common sources of inconsistency in sleep training:
Different caregivers responding differently. This is probably the biggest problem. Parents who respond differently to their child at night undermine training completely. One parent handles bedtime or night wakings differently than the other. If the parents are separated, there may be different rules in different homes. A grandparent who takes the baby overnight uses a completely different approach. The baby receives the slot machine — sometimes the lever produces the old response, sometimes it does not, and they optimise accordingly.
Every person who responds to your baby at night must be using the same approach, with the same response protocol, executed the same way. This is not a negotiation. If one caregiver cannot commit to the method, they should not be on night duty during training.
Caving on hard nights. Night two of cry it out is statistically the hardest night. Many parents who abandoned training and concluded it does not work for their baby quit on night two — which is the night before the turning point. Night three typically shows a dramatic reduction. Night two caving teaches the baby that two nights of sustained crying produces the original response, which means next time they need to sustain it for two nights to get there. Each inconsistent caving raises the bar for what gets a response, which makes the next attempt harder.
Responding differently to night wakings versus bedtime. Some parents do the method perfectly at bedtime but respond differently to a 2 a.m. waking. The baby does not distinguish bedtime from 2 a.m. — both are opportunities to settle independently, and the behaviour that works at 2 a.m. is the behaviour they will try at bedtime.
The fix: Before your start date, write down your response protocol in specific enough detail that either caregiver can execute it correctly at 3 a.m. without discussion. Then have an explicit conversation with every person who will be responding to your baby at night, agree on the protocol, and agree that no deviations happen for a minimum of fourteen days. If someone cannot commit to this, make a different plan for their nights during training.
Mistake 3: Training Nights But Not Naps
This mistake is common because it feels logical in the moment. Nights are the urgent problem. Days feel manageable. So parents apply their chosen method at bedtime and overnight, and continue rocking or feeding their baby to sleep for naps during the day.
This is not a half-measure. It actively makes nap training harder when you get there.
Sleep training only for nights without training for naps at the same time is not advisable. After a few days of night-only training, the child might sleep well at night. However, when parents try to later tackle nap sleep, the child uses all of their stamina from solid night sleep to fight their naps. Making nap training extremely difficult.
The underlying biology explains why. At bedtime, melatonin levels in your baby’s body are naturally elevated — the circadian clock is actively driving them toward sleep. This biological assistance means that even a baby who still has a strong sleep prop association will often fall asleep relatively quickly at bedtime because their body is working with them. During the day, melatonin is low, and sleep is driven primarily by accumulated sleep pressure — how long they have been awake. When a baby has consolidated their night sleep through training, they have more reserves during the day, which means they can sustain resistance to nap settling for longer.
The result of night-only training followed by delayed nap training is a baby who sleeps beautifully at night for a week, then fights naps ferociously when you finally address them — and parents who cannot understand why the same method that worked at night is failing during the day.
The solution is to apply your chosen method consistently to both naps and nights from day one. This does not mean using the same interval schedule at nap time as overnight — nap training has some differences — but the fundamental principle is the same: your baby should be put down awake and expected to settle independently for both sleep periods.
Nap training is genuinely harder than night training for the melatonin reasons described above. Naps may take longer to consolidate and may involve more protest than nights. This is normal. Hold the method consistently across both sleep periods and both will resolve.
The fix: When you build your sleep training plan, explicitly include nap training alongside night training. Write down your nap protocol the same way you write down your night protocol. Apply it consistently from day one.
Mistake 4: Skipping the Environment and Schedule Setup
Most parents who fail at sleep training skip two preparation steps that the research consistently shows are prerequisites for success: setting up the sleep environment correctly, and establishing a consistent bedtime routine before the training method begins.
On the environment: the sleep space needs to be correctly configured before any method can work at its full potential. Dark room — not dim, dark. White noise running continuously through the night, not just at bedtime. Room temperature between 68 and 72 degrees Fahrenheit. A safe crib with no loose items. These are not optional add-ons. A room with light leaking in at 5 a.m. is fighting against the method every single morning. White noise that cuts off at midnight is not doing the job it needs to do during the lightest sleep phases of the early morning. Get this right before night one, not during it.
On the schedule: if your baby is going to bed too late or not getting enough sleep during the day, your baby may wake up as often if not more often while sleep training and after. An overtired baby is a harder-to-train baby. Overtired babies produce more stress hormones, making them more difficult to soothe. They cry more intensely and become more overtired — a feedback loop that makes any method look like it is failing even when the method itself is being applied correctly.
This is one of the most demoralising experiences in sleep training: you are doing everything right, the method is correct, you are being consistent, and your baby is still waking frequently and crying more than expected — because the wake windows are wrong, the nap schedule is producing cumulative overtiredness, and you are trying to train a baby who arrives at bedtime already in a cortisol spike. The fix is not a better method. The fix is getting the schedule right before you start.
The bedtime routine is equally non-negotiable. A bedtime routine should be in place and consistent for at least one week before you begin your chosen method. The routine is what signals to your baby’s brain that sleep is coming — it is the sequence of cues that activates the sleep system before you ever put them in the crib. Without it, the method starts from a cold baseline every night. With it, your baby arrives at the crib already primed for sleep, which makes settling faster and crying shorter.
The fix: Do not start your method until three conditions are met: the sleep environment is correctly set up and has been used consistently for at least three nights, the bedtime routine has been established and is consistent, and the nap schedule is age-appropriate with no chronic overtiredness. This preparation week feels slow when you are exhausted and desperate. It makes the training itself dramatically faster.
Mistake 5: Quitting During the Extinction Burst
This is the mistake that turns a sleep training success into a sleep training failure, and it happens to parents who are doing everything else correctly.
The extinction burst is a well-documented behavioural phenomenon. When a previously reinforced behaviour — in this case, crying until a parent comes — is no longer being reinforced, the behaviour does not immediately stop. It first escalates. The brain, recognising that the usual response is not coming, increases the intensity and duration of the behaviour in an attempt to produce the result. This is the extinction burst: a spike in crying that typically occurs around nights three or four, just when parents are beginning to see progress and expecting continued improvement.
Parents often feel guilty about sleep training. Often, if they had persisted, their child would have started falling asleep more easily with just a day or two more of effort. The extinction burst is the exact moment most parents quit. Night one was hard. Night two was harder. Night three showed real improvement — less crying, faster settling. Night four produced a spike back to night-one levels or worse. The parent, who was expecting continued improvement, concludes that the method has stopped working or that their baby is suffering in a way that is not resolving. They go in. They comfort. They provide the old response.
And in doing so, they teach their baby something very specific and very damaging: that four nights of sustained crying followed by an intense escalation on night four is what produces the response. The bar for future training has just been raised significantly.
This is not a parent doing something wrong. It is a parent doing something completely understandable without the information they needed going in. The extinction burst is entirely predictable — it happens in most sleep training attempts, it is well-documented in the behavioural science literature, and it resolves on its own within one to two nights if the method is held. The problem is not the burst. The problem is that most sleep training guides do not tell parents it is coming.
How do you know the difference between the extinction burst — which you should hold through — and a genuine signal to stop? The burst is a spike in the same type of protest crying that has been present throughout training. It does not include a change in cry quality — no signs of pain, illness, or genuine distress beyond what has been present. It typically occurs in the night three to five window. It resolves within one to two nights of consistent holding.
Stop and reassess if: the cry quality changes significantly and sounds like pain rather than protest, your baby develops a fever or other signs of illness, or you are more than two weeks into consistent training with zero measurable improvement across any night.
Everything else is the extinction burst. Hold your protocol.
The fix: Before your start date, read everything you can about the extinction burst and make a specific plan for how you will handle it. Write it into your training plan: “If crying spikes significantly around nights three to five, this is the extinction burst. We will not change our response. We will hold the protocol for two more nights before reassessing.” Having this written down before you are in the middle of it is the difference between holding and caving.
What to Do If You Have Already Made These Mistakes
If you have read this list and recognised your previous attempt in it — you started too early, you were inconsistent, you quit on night four, you did not address naps — the most important thing to know is that none of this is permanent damage.
Babies are not trained into permanent behaviours by failed sleep training attempts. Their sleep associations reset relatively quickly, and a well-executed attempt at the right time, with the right preparation, will work regardless of what previous attempts looked like. The slate is not clean — a baby who has been taught through inconsistency that sustained crying eventually produces a response will apply that learning to the new attempt — but it is not irrevocably marked either.
The reset process: take one week to implement the preparation steps — sleep environment, bedtime routine, schedule audit — before starting again. Choose a method and commit to it with every caregiver. Write the protocol down. Identify your extinction burst plan before you start. Set a fourteen-day window for evaluation and do not change the method within that window except for genuine medical reasons.
If you have done all of this consistently for fourteen days and are not seeing meaningful improvement, it is time to bring your pediatrician in. Sleep training works for the vast majority of healthy babies with typical sleep associations when applied correctly. Persistent failure after genuine, consistent effort is a signal worth investigating medically — not a reflection on the method or on you as a parent.
The five mistakes on this list are not character flaws. They are information gaps — the things nobody told you before you started. Now you have the information. The next attempt will be different.
Sources: Dr. Craig Canapari, MD, Yale Pediatric Sleep Center, Top Ten Sleep Training Mistakes (updated October 2024); UChicago Medicine, Parent’s Guide to Sleep Training (2023); PedsCases, Sleep Training — Developed by Francis and Woods for PedsCases.com (November 2023); Little Dreamers, Six Common Mistakes Parents Make When Sleep Training (2022); Baby Sleep Trainer, 7 Common Baby Sleep Training Mistakes (2021); Scientific American, Does Sleep Training Work? (May 2025); Woolino, 8 Most Common Baby Sleep Training Mistakes (2022)
This article is for informational purposes only and does not constitute medical advice. Always consult your pediatrician before beginning any sleep training program, especially if your baby has underlying health conditions.