Baby Sleep Regressions and Sleep Training Methods Compared

Sleep regressions are temporary, not a diagnosis. They cluster around 4, 8–10 and 18 months because those ages are developmental leaps, not because you "broke" sleep. If you choose to sleep train, methods range from no-cry (chair, fading) to faster cry-it-out variants (Ferber, extinction). None is medically required. Safer-sleep rules do not change: back, clear cot, room-share for six months.

Who should skip this sleep-regression guide?

Skip this if your baby is under 12 weeks — that is newborn biology, not a regression. Use the newborn sleep guide instead. Skip it if a paediatrician or health visitor has given you a medical sleep plan (reflux treatment, nasogastric feeds, a NICU discharge routine). Skip sleep training advice entirely if you are in the middle of an illness, a vaccination week, or a house move — wait until the baseline is boring again.

Why do sleep regressions happen, and when?

They happen because sleep architecture and the rest of development do not move in a straight line. A baby who was giving you a 6-hour stretch can suddenly wake every 90 minutes. That is usually a leap, not a failure of your routine.

The NHS page on helping your baby to sleep is the source for "no one method is required" and for keeping nights boring and dark (NHS — helping your baby to sleep, checked 12 August 2026). The Lullaby Trust is the source for the safer-sleep list that still applies during a regression (Lullaby Trust, checked 12 August 2026).

How do sleep training methods compare?

Sleep training is a personal choice. Plenty of families never do it. If you want a structured approach, the honest difference is speed versus how much crying you are willing to hear, not moral virtue.

• Chair method: you sit in the room and move the chair further from the cot over nights. Slow (often 1–2 weeks). No leaving them to cry alone.
• Pick-up-put-down: pick up to soothe, put down when calm, repeat. Physically exhausting. No extended crying alone.
• Fading / gradual retreat: less rocking, then in-room, then out, over a week or two.
• Ferber (graduated extinction): leave, return at lengthening intervals to reassure without picking up.
• Extinction ("cry it out"): settle, leave, return only if you are worried about safety. Fastest (often 3–7 nights) and the hardest for many parents.

None of these is an NHS requirement. If a method needs you to ignore a fever, a cough, or a baby who is hard to wake, stop — that is health, not sleep training.

Does safer sleep still apply if you sleep train?

Yes. Always. Back to sleep, clear cot, no pillows or duvets, room-share for the first six months, 16–20°C where you can. A "sleep trainer" pillow, a nest, or a monitor that claims to prevent SIDS is a red flag. See the sleeping bag guide for tog and fit, and the monitor guide if you want reassurance kit that does not pretend to be medical.

What helps sleep regardless of method?

A boring, repeatable wind-down (bath, feed, bag, dark room) beats any branded technique. Rule out teething, illness and reflux before you "train". A correct-tog sleeping bag removes a whole category of night wakes. White noise is optional; a white noise machine is a tool, not a cure. Consistency for a week tells you more than switching method every other night.

If you want one next action: keep the safer-sleep list tonight, pick one settling approach for seven days, and only then decide whether to change it.