The five non-stimulant options
Atomoxetine (Strattera)
- Mechanism: Selective norepinephrine reuptake inhibitor
- Duration: 24-hour coverage
- Onset: 4-8 weeks for full effect
- Best for: Adults who can’t take stimulants, anxiety + ADHD, substance use history
- Side effects: Nausea (much better taken with food), sleep changes, reduced appetite, sexual side effects in adults, urinary hesitancy; rare liver injury
- Effect size: Moderate
- Watch: Metabolised by CYP2D6 — poor metabolisers and anyone on fluoxetine or paroxetine can get much higher levels from a standard dose
Atomoxetine was the first non-stimulant approved for ADHD (2002) and is now generic and cheap, which makes it the one most people are offered first. Its defining problem is the wait: four to eight weeks before you know whether it works, which is a brutal ask of an ADHD nervous system and the reason most abandoned trials are abandoned. See our full Strattera guide for dosing, the side effects adults are rarely warned about, alcohol, and the honest Strattera-vs-Qelbree comparison.
Viloxazine (Qelbree)
- Mechanism: Norepinephrine reuptake inhibition plus serotonergic activity (5-HT2B antagonist, 5-HT2C agonist)
- Duration: Once daily, 24-hour coverage
- Onset: Faster than the others — separation from placebo as early as week 1–2
- Best for: Adults who need a non-stimulant but couldn’t survive atomoxetine’s six-week wait; prominent irritability
- Side effects: Somnolence, fatigue, reduced appetite, nausea, insomnia, irritability; boxed warning for suicidal thoughts and behaviours
- Watch: Strong CYP1A2 inhibitor — it slows caffeine clearance, so your usual coffee intake can suddenly feel like double
Qelbree is the newest of these — FDA-approved in 2021 for ages 6–17 and in 2022 for adults, and the first genuinely new non-stimulant mechanism to reach the ADHD shelf in roughly two decades. It matters most for the people atomoxetine has failed on timing rather than on effect: a two-week trial is one an ADHD adult can actually complete, where a six-week trial in the dark is one they abandon. It is also the option whose interactions are least often explained — see our full Qelbree guide for dosing, the boxed warning, the caffeine problem, and the honest Qelbree-vs-Strattera comparison.
Guanfacine (Intuniv)
- Mechanism: Alpha-2 adrenergic agonist
- Duration: 24-hour coverage
- Onset: Weeks for full effect
- Best for: RSD, emotional dysregulation, ADHD + tics, ADHD + anxiety, augmenting stimulants
- Side effects: Sedation, hypotension, dizziness, fatigue
- Effect size: Moderate; particularly strong for emotional features
Clonidine (Kapvay)
- Mechanism: Alpha-2 adrenergic agonist (similar to guanfacine but different receptor subtypes)
- Duration: Variable; immediate and extended release
- Best for: Sleep difficulties, hyperactive ADHD, tic disorders, often as add-on
- Side effects: Sedation, hypotension, dizziness
- Effect size: Moderate
Bupropion (Wellbutrin)
- Mechanism: Norepinephrine and dopamine reuptake inhibitor
- Status: Off-label for ADHD (FDA-approved for depression and smoking cessation)
- Best for: ADHD + depression, ADHD + nicotine cessation, adults who want stimulant-like effect without controlled substance
- Side effects: Anxiety, insomnia, dry mouth, lower seizure threshold (contraindicated in some adults)
- Effect size: Smaller than stimulants but real
When non-stimulants are preferred
- History of substance use disorder
- Cardiac concerns making stimulants risky
- Anxiety amplification on stimulants
- Stimulant side effects intolerable
- Hypersensitive nervous system
- Need for 24-hour coverage without peaks and troughs
- Specific co-occurring conditions (tics, depression)
- Augmenting stimulants for fuller coverage
Combining stimulant + non-stimulant
Common practice. Most common combinations:
- Stimulant + guanfacine: Stimulant for attention, guanfacine for RSD and emotional regulation
- Stimulant + atomoxetine: Less common but used for fuller 24-hour coverage
- Stimulant + bupropion: Possible with cardiac monitoring; addresses depression alongside ADHD
Stimulant vs non-stimulant: the honest comparison
Start with the finding nobody should bury: on average, across populations, stimulants outperform non-stimulants on the core symptoms of ADHD. Stimulant effect sizes in adults sit in the region of 0.5–0.8; non-stimulants land lower, broadly around 0.4–0.5. That gap is real, it has been stable for a long time, and it is why stimulants are tried first for most adults who can take them.
But an average is not a person, and the comparison people actually face is not “which drug scores higher in a meta-analysis” — it is “which set of trade-offs can I live with for years.” On that question, non-stimulants win several categories outright:
- No peak means no crash. Nothing arrives, so nothing departs. There is no 4pm cliff, no rebound irritability while the dose leaves, no evening where you become a worse version of yourself because the medication is wearing off.
- Coverage is flat and continuous.Stimulants cover a working day. Non-stimulants are simply present — including at 7am with the children, and at 9pm when the emotional regulation you needed all day has run out. For adults whose hardest hours are the ones a stimulant does not reach, this is not a minor detail.
- No controlled-substance apparatus. None of the five is scheduled. No monthly prescription that cannot be refilled, no pharmacy shortage roulette, no justifying yourself at a counter. That is an executive-function tax removed from the people least able to pay it.
- Nothing to misuse. For adults in recovery, this is frequently the difference between being medicated and not.
- Appetite and sleep are usually less brutalised— though guanfacine and clonidine bring sedation of their own, and atomoxetine and viloxazine can suppress appetite too. Different costs, not zero costs.
And the framing that gets lost in the versus: this is not a binary. Combining a stimulant with a non-stimulant is common, well-supported practice, precisely because they cover different dimensions — the stimulant handling attention and the non-stimulant handling the emotional regulation, the evenings, and the edges the stimulant never reached. If you have been treating this as a fork in the road, you may be asking a narrower question than the one actually available to you.
The slower onset reality
Stimulants work acutely — you feel it within hours. Non-stimulants build up over weeks. This requires patience during initial trial:
- Atomoxetine: 4-8 weeks for full effect
- Viloxazine (Qelbree): faster — separation from placebo as early as week 1–2
- Guanfacine: weeks for full effect, more gradual onset
- Bupropion: 2-4 weeks typically
The slower onset means dose-finding takes longer. But also no rebound, no peak-trough variability, smoother experience day to day.
There is a trap hiding in those weeks, and it is worth naming because it is the single most common way a non-stimulant trial fails. You are asking an ADHD adult — someone whose nervous system is specifically poor at sustaining an unrewarding effort towards a delayed and uncertain payoff — to take a daily pill that does nothing perceptible, through a fortnight of side effects, on the promise that something might happen later. That is close to a worst-case executive-function task. People abandon medications that would have worked for them, not because the drug failed, but because the waiting did.
Two things help. Decide in advance how long the trial runs, write it down somewhere you will actually see it in week two, and treat that date as the decision point rather than re-litigating it every morning. And do not rely on your own perception alone: the effect of a non-stimulant is usually easier to see from the outside than the inside, so ask the people who live with you whether anything has changed. “Can I feel it?” is the wrong question. “Is my week going better than it was?” is the right one.
Side effect comparisons
Sedation
Guanfacine and clonidine often produce sedation, particularly initial doses. Sometimes useful for sleep. Atomoxetine and bupropion generally not sedating.
Anxiety
Bupropion can increase anxiety. Atomoxetine sometimes affects mood. Guanfacine generally reduces anxiety. Choosing based on anxiety baseline matters.
Sleep
Atomoxetine can affect sleep variably. Bupropion often worsens sleep. Guanfacine and clonidine often improve sleep. Timing of dose matters.
Cardiac
Guanfacine and clonidine lower blood pressure. Bupropion can raise it. Atomoxetine can raise heart rate. All require some cardiac monitoring in adults with cardiac history.
How prescribers choose
- RSD or emotional dysregulation prominent → guanfacine
- Anxiety + ADHD → atomoxetine or guanfacine
- Depression + ADHD → bupropion
- Substance use history → non-stimulant generally
- Sleep difficulty + ADHD → guanfacine or clonidine
- Tics + ADHD → guanfacine or clonidine
- Want stimulant-like effect without controlled status → bupropion
FAQ
What non-stimulant ADHD medications exist?
Five main options: atomoxetine (Strattera), viloxazine (Qelbree — FDA-approved for adults in 2022, the newest of the group), guanfacine (Intuniv), clonidine (Kapvay), and bupropion (Wellbutrin — off-label for ADHD). Each works on different neurotransmitter systems and has a different profile. Generally smaller effect than stimulants, but none is a controlled substance, none has meaningful abuse potential, and all are useful for adults who can’t take or tolerate stimulants.
Who needs non-stimulant medication?
Adults for whom stimulants don’t work or aren’t appropriate. Reasons: history of substance use disorder making stimulants risky, cardiac concerns, anxiety amplification on stimulants, side effects intolerable on stimulants, hypersensitive nervous system, augmenting stimulants for fuller coverage, specific co-occurring conditions where non-stimulant has dual benefit (guanfacine + tics, bupropion + depression).
Are non-stimulants weaker than stimulants?
Generally smaller effect sizes. In adults, stimulants run roughly 0.5–0.8; non-stimulants around 0.45–0.5 (moderate). But: smaller doesn’t mean ineffective. Many adults respond well to non-stimulants. And the smaller effect comes with different benefits — sustained 24-hour coverage, no rebound, no controlled substance status, often complementary side effect profile.
How do non-stimulants work differently?
Atomoxetine: blocks norepinephrine reuptake. Slow build-up (weeks). 24-hour coverage. Guanfacine and clonidine: alpha-2 adrenergic agonists. Affect norepinephrine signalling differently. Often help with emotional regulation and sleep. Bupropion: norepinephrine and dopamine reuptake inhibitor. Closest mechanistically to stimulants among non-stimulants. Different mechanisms produce different clinical effects.
How long do non-stimulants take to work?
Atomoxetine: 4-8 weeks for full effect. Patience needed. Viloxazine (Qelbree): the fastest of the group — improvement separated from placebo as early as week 1-2 in trials. Guanfacine: weeks for full effect, more gradual onset. Bupropion: 2-4 weeks typically. All are different from stimulants (acute effect within hours). The slower onset means a longer dose-finding process but also no rebound and no peak-trough variability. The waiting is where most non-stimulant trials fail — decide your trial length in advance and judge it on whether your week went better, not on whether you can feel the pill.
Can I take stimulant and non-stimulant together?
Yes, common practice. Guanfacine + stimulant is particularly common — guanfacine addresses RSD and emotional regulation; stimulant addresses attention. Atomoxetine + stimulant less common but used. Bupropion + stimulant possible but cardiac considerations. Combination therapy expanded the options for adults who don’t get full benefit from monotherapy.
What about side effects?
Variable by medication. Atomoxetine: nausea, sleep changes, sometimes mood effects, rare hepatic effects. Guanfacine: sedation, hypotension, dizziness, fatigue. Clonidine: similar to guanfacine. Bupropion: anxiety, insomnia, dry mouth, lower seizure threshold (contraindicated in some adults). Generally well-tolerated but trial-and-error to find the right fit.
Are non-stimulants right for women in pregnancy?
Limited pregnancy data for atomoxetine. Bupropion has more pregnancy data than atomoxetine, but it remains limited and mixed — not a reason to assume it is safer. Guanfacine and clonidine have specific pregnancy considerations (blood pressure effects). For specific medication-and-pregnancy decisions, see our adhd-medication-pregnancy guide and discuss with prescriber. Pregnancy decisions are individual.