Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the child ADHD research overview.
Short answer. Most parents who say "we've been consistent for years and nothing works" are running effective behavioural strategies, but without the measurement layer needed to detect the slow frequency shifts the literature predicts. The MTA Cooperative Group (1999) and Pelham & Fabiano (2008) document that behavioural parent training produces meaningful, durable change — but the change shows up as gradual reductions in event frequency over months and years, not visible week-over-week transformation. A genuine multi-year plateau exists, but it is rarer than it feels, and the research-backed audit below separates the two before any strategy is abandoned.
The MTA Cooperative Group's 14-month randomized trial (1999) and its multi-year follow-ups remain the largest controlled evidence base in child psychiatry. The headline finding is well known: combined medication and behavioural treatment outperforms either alone for functional outcomes. The less-cited finding matters more for parents who feel stuck: even within the active-treatment arms, gains were uneven, partial, and substantially better detected by structured ratings than by parental impression. Pelham & Fabiano's (2008) review of behavioural interventions reports moderate-to-large effect sizes — but emphasises that the effects accumulate slowly and require sustained fidelity.
Barkley's longitudinal work on ADHD into adolescence and adulthood reframes the question further: ADHD is a chronic, developmentally evolving condition, and the right outcome metric is not "is the child fixed" but "is the child's trajectory better than it would have been without the intervention." That counterfactual is genuinely invisible to the parent in the room. The household can be on a substantially better trajectory than the no-treatment counterfactual and still feel, day to day, like nothing has changed.
This is the central asymmetry: the research is measuring 14-month, 3-year, and 8-year functional outcomes against a control group. The parent is measuring this morning against last Tuesday.
When parents say "consistent for years and nothing works," they almost always mean one of three things:
1. A specific behaviour that has not visibly improved. Morning meltdowns. Homework wars. Sibling aggression. The frequency feels identical to two years ago. 2. A felt sense of exhaustion. The parent is running the techniques but no longer believes in them. "I'm starting to feel like I'm experimenting on my own child" is a common phrasing on parent forums. 3. A specific recent regression. A bad month is being read as evidence that the whole multi-year effort has failed.
Each of these has a different research-backed answer, and none of them can be diagnosed from memory.
Pelham & Fabiano (2008) are explicit that behavioural change shows up as frequency shifts in target behaviours, not as the disappearance of those behaviours. A child who had four homework meltdowns a week and now has two is a 50% improvement — and it will not feel like one, because the meltdowns that remain are fully as intense as the ones that did not. "It still feels exactly the same" is the predicted subjective experience of a 50% reduction, because human memory weights the worst recent event, not the count.
The audit: pick three concrete markers (e.g. meltdowns per week, morning prompts before leaving the house, minutes of homework before the first refusal). Ask honestly whether you have data on these from one year ago, two years ago, three years ago. Most parents who feel "nothing works" do not — the comparison is being made between today and an idealised memory of a year ago, which is not a real comparison.
The single biggest predictor of behavioural parent training outcome across randomised trials is parent fidelity (Kazdin, 2005; Pelham & Fabiano, 2008). A simple plan applied 90% of the time outperforms a sophisticated plan applied 50% of the time. Multi-year ADHD households almost always experience drift: the praise ratio that started at 4:1 has slid to 1:1 under chronic stress; the consequence delivered with a neutral tone two years ago is now delivered with frustration; planned ignoring has become unplanned avoidance.
Drift is not failure of the strategy. It is failure of the running conditions for the strategy. The audit is whether what you are currently running is what the literature actually evaluates, or a depleted version of it.
Children with ADHD have high rates of co-occurring anxiety, learning disabilities, oppositional patterns, and sleep disruption. A behavioural plan calibrated for ADHD that is now running into untreated anxiety or an undiagnosed learning disability will look like the plan stopped working — when in fact a new variable has entered. The AAP guideline (Wolraich et al., 2019) explicitly notes that comorbidity changes the treatment plan. "Consistent for years and now nothing works" sometimes means a comorbidity has emerged that the original plan was not designed for.
Children grow. Doses that were correct at age 7 are frequently subtherapeutic by age 10. The MTA (1999) follow-ups document that benefit attenuated in part because medication management drifted post-trial. "Nothing works any more" sometimes means "the dose has not been re-evaluated against the current child." This is a separate question from whether to switch medications, covered in when should I ask to switch my child's ADHD medication?.
1. Run the four audits above before concluding the strategy has failed. Most "nothing works after years" verdicts dissolve at audit 1 (no real frequency data) or audit 2 (drift in fidelity). 2. Establish weekly markers and run them for a 90-day window before making any major change. Pelham & Fabiano's effect sizes are detected over 8–12 weeks of consistent measurement, not from impressions. 3. If a comorbidity audit raises a flag, get it assessed before changing the ADHD plan. Adding behavioural firepower to an untreated anxiety or sleep problem is rarely productive. 4. If medication is part of the picture, request a structured medication review with weekly data — not memory. The AACAP and AAP guidelines both treat medication titration as a data-driven process.
The parents in the longitudinal literature whose children had the best outcomes were not the ones running the cleverest strategies. They were the ones running an evidence-based strategy long enough, and consistently enough, to let the effects accumulate (MTA Cooperative Group, 1999). Years of consistency, audited honestly, is almost always closer to working than it feels.
---
Unseen Progress publishes long-form caregiver research. See the full child ADHD research overview for the complete framework.