what is a dream feed and how do I do it?

A dream feed is a scheduled feed given to a sleeping baby at around 10-11pm, before you go to bed. Lift baby without fully waking them, offer breast or bottle and let them feed drowsily, then return them to the crib still sleepy — no nappy change unless there has been a bowel movement. The goal is to top up their calories so they don't wake hungry at midnight or 1am.

do dream feeds actually help babies sleep longer?

Sometimes. The evidence shows dream feeds as part of broader sleep interventions (Paul 2011, Stremler 2013) produce modest improvement — but dream feeds have never been tested in isolation in a large RCT. They work when hunger is genuinely driving the early night wake-up, but have no effect if the baby wakes due to a sleep association rather than hunger.

how long should I try a dream feed before stopping?

Give it 10-14 nights. If the first natural wake-up hasn't shifted at least 1 hour later, the dream feed is not working for your baby and continuing is just an extra disruption to your evening. Signs it is working: the wake-up shifts later, baby resettles easily after the dream feed.

how do I do a dream feed without waking the baby fully?

Keep lights off or use dim red light only. Lift baby while still swaddled if they are in the swaddle phase. Rouse them just enough to latch or take the bottle — a little stirring is fine, fully waking them is counterproductive. Have everything (breast accessible or bottle pre-warmed) ready before lifting. Skip the nappy change unless absolutely necessary.

can you dream feed a baby with reflux?

With extra caution. Laying a refluxing baby flat immediately after a feed can worsen symptoms. If your baby has reflux, hold them upright for 15-20 minutes after the dream feed before returning to the crib. For babies with significant reflux, dream feeds may cause more problems than they solve and are worth discussing with your pediatrician first.

Dream Feeds: Do They Actually Work? | Sleep Through The Night

The idea is appealing: sneak in at 10pm, offer a feed to your already-sleeping baby, and buy yourself a longer stretch of uninterrupted sleep before the inevitable 2am wake-up. Some parents swear by it. Others try it for two weeks and see no difference whatsoever.

So what does the research say? The honest answer: it’s complicated, the evidence is limited, and results are genuinely individual.

What Is a Dream Feed?

A dream feed is a scheduled feed given to a sleeping or semi-sleeping baby, typically between 10pm and 11pm — before the parents go to sleep for the night.

The goal: top up baby’s tank so they don’t wake hungry at midnight or 1am, shifting the first night waking to 2–3am or later.

The technique: you lift baby gently without fully waking them, offer the breast or bottle, let them feed drowsily, then return them to the crib still sleepy.

What the Research Shows

Paul 2011

Paul et al. studied a structured intervention combining a scheduled dream feed with graduated extinction for infants 6–8 weeks old. The study found that the combined intervention produced longer sleep stretches compared to control, with the dream feed being a component of the bedtime routine that appeared to support the overall behavioral strategy.

Caveat: the dream feed was part of a package of interventions, not isolated as a standalone variable. It’s difficult to attribute the benefit specifically to the dream feed versus the behavioral sleep training component.

Stremler 2013

Stremler et al. conducted a randomized controlled trial of a behavioral sleep intervention for mothers with infant sleep problems. The intervention included dream feeds as one element of a multi-component approach (alongside consistent bedtime routines and parental education). Results showed modest but statistically significant improvement in infant sleep and maternal fatigue compared to control.

Again: dream feeds were bundled with other strategies, not tested in isolation.

The Honest Assessment

There are no large, well-designed RCTs that specifically isolate dream feeds as the independent variable. The existing evidence shows dream feeds as part of multi-component behavioral interventions that work — but we can’t cleanly attribute how much of the benefit comes from the dream feed itself versus the other components.

What this means practically: dream feeds might help, and they’re unlikely to cause harm if done correctly. But if you try them and see no improvement after 1–2 weeks, the evidence doesn’t suggest you’re doing it wrong — it may simply not work for your baby.

Why It Works for Some Babies and Not Others

Sleep extension from a dream feed depends on:

Whether hunger is actually driving the wake-up. If your baby wakes at 1am because they’re hungry, a dream feed at 10:30pm may extend that to 2:30–3am. If they wake at 1am due to a sleep association (needing to be rocked or nursed back to sleep rather than hunger), a dream feed won’t address the root cause.

The baby’s gastric capacity. Younger, smaller babies have small stomachs and genuinely need more frequent feeds regardless of when the last one was. A dream feed doesn’t change biology.

Individual variation. Some babies absorb a sleepy feed efficiently; others take in very little when not fully awake and end up waking at the same time anyway.

How to Do It

Timing

10pm–11pm is the typical window. Going later (11:30pm+) may interfere with your own sleep and risks waking baby too thoroughly.

Technique

Lift baby gently from the crib without turning on bright lights
Keep them swaddled if they’re still in the swaddle phase
Rouse them just enough to latch or take the bottle — a little stirring is fine, fully waking them is counterproductive
Offer breast or bottle; let them feed at their own pace
Wind/burp gently (see reflux caveat below)
Return to crib before they fully wake up

Don’t change the diaper

Unless there’s been a bowel movement, skip the diaper change. A nappy change is enough stimulation to fully wake a baby you’ve worked to keep drowsy. A wet nappy won’t harm them overnight.

For breastfeeding parents

Have the breast ready before lifting. Fumbling with clothing while holding a stirring baby tends to fully wake them. Side-lying or semi-reclined positions can work well.

For bottle-feeding parents

Pre-warm the bottle before you go to bed so it’s ready at 10pm. A cold bottle will wake them up fast.

When to Stop Trying

Give it 10–14 nights. If you’re seeing no change in the timing of the first wake-up, the dream feed isn’t working for this baby and continuing is just an extra disruption to your evening.

Signs it’s not working:

Baby still wakes at the same time as before the dream feed
Baby is harder to settle after the dream feed than they were before
Baby starts waking more often, not less (sometimes adding a feed teaches them to expect it)

Signs it’s working:

First natural wake-up shifts later by 1+ hours
Baby is easy to re-settle after the dream feed and returns to sleep quickly

The Reflux Caveat

If your baby has gastroesophageal reflux (GERD), dream feeds require extra caution.

Lying a baby down immediately after feeding (which is what the dream feed technique involves) can worsen reflux symptoms — stomach contents move back toward the esophagus more easily when supine. If your baby refluxes, consider:

Holding them upright for 15–20 minutes after the dream feed before returning to the crib
Using a slight head elevation on the mattress (check with your pediatrician — loose mattress elevations are a safety hazard; purpose-built tilts exist)
Discussing with your doctor whether dream feeds are appropriate given the severity of their reflux

For babies with significant reflux, dream feeds may cause more problems than they solve.

The Bottom Line

Dream feeds are low-risk, moderately evidence-supported as part of a broader sleep strategy, and worth trying if you’re willing to give them 1–2 weeks. They work for some babies and have no effect on others — and the research doesn’t yet give us good predictors of which camp your baby will fall into.

If it works: great. You’ve bought yourself a longer first stretch.
If it doesn’t: stop. It’s not failure — it’s information about what your baby actually needs.

Either way, you’re one data point closer to a plan that does work.