7 potty training regressions: when progress went backwards

The third accident of the morning came at 11:40am. I was already drafting the speech about starting over. A potty training regression is a stretch where a child who was using the potty stops. The accidents come back, and the panic comes with them. We counted seven of those stretches between May and August; three of them were not regressions at all. Here is every one in order, with what we did and how long it lasted.

Here is the whole list at a glance, then each one in detail.

1. The false start: the bootcamp quit

The first backward step happened before training started. In May, on the Whitsun long weekend, I decided we were doing this; the partner did not share the urgency. I had read half of Oh Crap, Potty Training. The plain white potty moved into the living room; the plan was three focused days. Day two ended it. At 7:10pm she had an accident on the living-room rug. She looked at the puddle, then at me; I saw the project through her eyes. A potty in the middle of the room, a parent who had turned a bodily function into a deadline. We quit that evening, and it felt like losing. It was the first correct call, because you cannot regress from a place you were dragged to. The no-pressure timeline we kept replaced the bootcamp a week later.

2. The new-baby lever: the attention regression

The classic trigger got a new stage in our house. The potty became a lever. Our first month with two showed the move in its first form, before the potty existed. At 4:03pm on a Thursday, while the baby latched, the toddler announced a toilet emergency. We rushed her to the bathroom and sat her down. Nothing happened. She did not need the toilet; she needed the feed to stop. The potty gave the lever a better excuse than the one she had used in spring. The tell was the timing: the emergency always evaporated the moment the feed ended. The fix was not potty strategy; it was attention given before it had to be demanded. Ten minutes of one-on-one a day, given early, is the whole trick. Once the attention arrived on its own, the emergencies lost their audience; the lever went back to being a potty. The failure mode ran the other way. Answering every announcement feeds the lever; ignoring them all drowns the real signal.

3. The wet-nap scare: the regression that was the plan

The no-pressure plan kept naps and nights in nappies; daytime was the project. A wet nap was not an accident. It was the plan doing its job, which my panic did not care to hear. One Tuesday in week four she woke from her nap wet for the first time in days. I stood in the doorway doing the arithmetic again, then I read our own notes. Dry naps only became common by week five; we were in week four. The plan had not failed; I had forgotten the plan. Expectation is a harsh referee; it calls fouls on plays that were never in the game.

4. The structure collapse: the Ferien drift

Potty training ran on the kita clock without anyone noticing; the day did the asking: wake, meals, bath, bed. Then the two closed Ferien weeks arrived; the day lost its shape. Thursday of the second week, at 5:40pm, dinner started without the before-dinner trip. She stood up mid-meal; the floor did the rest. The accidents crept back. She had not forgotten how; nobody was asking. That part annoyed me most: the structure had been doing the training while I took the credit. The fix was boring. We put the trips back and stopped blaming the toddler for a schedule we had let go. Most toddler chaos is a scheduling problem; this was a textbook case.

5. The illness slide: the HFMD week

A sick child does not care about dry. Early August brought the hand-foot-mouth week from our first summer of kita illnesses: fever one evening, spots by morning. Five sick days followed, kita out. We did not train that week; we did not even try. The accidents returned anyway, because a body fighting a virus has no attention budget left for holding. I seriously considered re-training her on day two of the fever, which would have been a mistake. Waiting was the move. The trips worked again within days of the spots fading, like nothing had happened. Illness does not erase progress. It pauses it; the pause is not a loss.

6. The three-accident morning: the low point

One morning braided all of August’s strands together: the Ferien hangover, the illness tail, the baby’s fourth month. Three accidents before lunch. The third one I handled the prescribed way: changed, cleaned, no commentary, no face. Inside, I was drafting a speech about starting over that I never got to give. The partner stopped me with the actual fix. Back to transitions only, no reaction, and the log, she said; not a new plan, the old one. We asked at the anchors again. The slide faded within two weeks, which is how long most of these took. The low point taught me the shape of the fix: smaller, not bigger. Every instinct said escalate; the data said repeat the boring part. Here is the checklist for a morning like that:

• Ask at the anchors: wake, meals, bath, bed. Asked once.
• Say nothing about the accidents. No face, no lecture, no sigh.
• One sentence on the way to the potty, said once.
• Log the accidents. The notebook outlasts the panic.
• Give attention before it is demanded, especially with a new baby in the house.

7. The night-nappy standoff: the regression that never came

Night dryness runs its own timeline; ours was not due. The plan from the start was daytime first, nights in nappies; a wet night was never a regression. It still scared me once, at 3:12am, when the nappy was wet after a run of dry nights. The phrase “here we go again” crossed my mind before the calendar did. The plan says nights come later; the nappy was the plan and had done its job. We did not night-train her. One wet night did not start a new campaign. The standoff never happened. That was the point: the NHS potty training page trains daytime first; night dryness runs its own timeline.

How we picked these

• Every item is a real backward step from May to August 2026, in order.
• We kept a log where one existed: accidents per week from June to July, then running notes through the August regression.
• Every item had to be a real event we could date, with the trigger and the response attached.
• We kept the false alarms on purpose. They cost the most panic, which makes them the most useful.
• Regressions are a normal part of the toddler years; we logged the sleep version from the hallway the same way.

Filed under toddler. Everything we’ve written is in the archive. Our accuracy and sourcing rules live on the editorial policy page.