TL;DR
- •Roughly 50-70% of treatment-resistant patients respond to ketamine in clinical literature; the Murrough 2013 RCT documented 64% response within 24 hours of a single dose vs 28% placebo.
- •"Didn't work" and "didn't work yet" are different. Most response evaluations should happen after 4-6 sessions, not 1-2 — early non-response often converts with dose calibration, frequency adjustment, or integration support.
- •Predictors of better response: more severe and treatment-resistant baseline, motivated to engage with integration work, no active substance use, stable life circumstances (sleep, basic safety), and willingness to adjust other medications if needed.
- •Predictors of harder response: severe ongoing trauma without specialist support, active substance use disorder, severe untreated bipolar disorder, very long-standing depression with strong cognitive-belief structure, ongoing acute psychosocial stressors that overwhelm treatment.
- •When to adjust: dose, frequency, modality (sublingual vs IV vs esketamine), integration support. When to switch: 8-10 sessions with no movement, persistent intolerable side effects, deterioration during treatment, or active psychiatric instability that ketamine isn't the right tool for.
- •Non-response on ketamine doesn't close the door — TMS, ECT, IV ketamine, esketamine, psilocybin (in approved settings), trauma-specific therapies, and combination approaches all remain options.
Step by step
- 1
Sessions 1-2 — the early signal
About 30-40% of responders feel something within the first session — "lighter," "quieter," "more room to breathe." Others see nothing in session 1 and respond by session 3. Both patterns are normal. Don't conclude "ketamine doesn't work for me" after session 1 or 2; the data isn't in yet.
- 2
Sessions 3-4 — the meaningful evaluation point
By session 3-4, most patients who will respond have started to. PHQ-9 / GAD-7 tracking usually shows movement. Sleep, energy, emotional range, intrusive thinking are common change vectors. If you're not seeing change here, the conversation shifts to adjustment.
- 3
Sessions 4-6 — adjustment options if response is partial or absent
Dose increase or decrease (some patients need higher; some respond to lower); frequency change (sometimes more frequent sessions early build response); modality switch (sublingual to IV ketamine, or to esketamine via Spravato program); integration support (formal therapy added). The goal is to find the configuration that produces response, not to give up at the first protocol that doesn't.
- 4
Sessions 8-10 — the harder decision point
If 8-10 well-conducted sessions across adjusted protocols haven't produced meaningful response, the conversation shifts to whether ketamine is the right modality. This is the moment to honestly discuss switch options: TMS, ECT for severe cases, formal trauma-focused therapy (sometimes the underlying issue isn't a medication problem), or combination approaches.
- 5
Ongoing — sustaining response and recognizing wane
For responders: maintenance dosing sustains response. Watch for: response wearing off between sessions, life events overwhelming the treatment, new symptoms emerging. Mid-course adjustments are normal. The treatment is responsive, not fire-and-forget.
What this actually feels like
When ketamine is working, patients describe a specific phenomenology: depression becomes "quieter," "further away," "no longer the dominant voice" in their head. Energy returns. Emotional range opens — being able to cry, laugh, feel love fully. Sleep improves. Forward-looking thinking returns; small actions become possible. The change is usually gradual but accelerated compared to SSRIs — measured in sessions rather than months. When ketamine isn't working, patients often describe a sense that "nothing has shifted" after multiple sessions — not just mood but the broader pattern of depression. Sometimes the medication produces a brief lift but doesn't consolidate; sometimes nothing at all. This is information, not failure, and points toward adjustment or alternative approaches.
Timeline
Early signal: hours-to-days after first session (in responders). Meaningful evaluation: 4-6 sessions over 2-3 weeks. Decision point on protocol: 8-10 sessions if response is absent or weak. Modality switch: typically after 8-10 sessions without meaningful response despite adjustment. Don't make the "is this working?" decision after 1-2 sessions — the data isn't in. Don't prolong an unsuccessful protocol past 10 sessions without honest discussion of switch options.
Common concerns, addressed
“I had session 1 and felt nothing — is this a failure?”
No. Many responders see nothing in session 1 and respond in sessions 2-4. The first-session non-response rate is meaningfully higher than the 4-6 session non-response rate. Don't conclude based on one data point. The evaluation point is session 3-4 minimum.
“I responded after session 3 but the effect is gone by next session”
Normal early pattern. Response duration starts at 3-7 days for many patients and builds with subsequent sessions. The induction phase is specifically about building durability. If durability isn't developing by session 6-8, frequency adjustment or dose change can help.
“How do I know if I'm one of the 30% who won't respond?”
You don't, until you've tried adequately. Most "non-responders" haven't tried 6-8 sessions with optimized protocol — they've tried 2-3 sessions of one configuration. Genuine non-response requires adequate testing: 8-10 sessions with dose adjustment, frequency optimization, and adequate integration support. The number who fail after that test is smaller than the number who fail after 2-3 sessions of a default protocol.
“My therapist thinks ketamine is masking the "real work" I need to do”
The medication-vs-therapy framing is often false. Many patients need both — ketamine to relieve symptoms enough to do the deeper work, therapy to address what surfaces in the response window. If your therapist is dismissive of medication, that's worth examining. If they're genuinely concerned about you avoiding meaningful psychological work, that's worth examining too. Often both are partially right.
“I responded but my life is still hard”
Common and meaningful. Ketamine treats depression / anxiety / PTSD symptoms — it doesn't resolve life difficulties. Many patients describe feeling "able to face" hard situations they were previously overwhelmed by — but the situations still need facing. This is response, not failure. Integration work (therapy, behavior change, life-context adjustments) is where the response gets translated into sustained life change.
“I want to stop ketamine because the benefits aren't worth it”
Legitimate question worth real conversation. Discuss with your physician honestly: what's working, what's not, what the alternatives are. Discontinuation is a real option; so is adjustment that addresses the specific issues. Patient autonomy matters; informed decisions matter. Don't feel obligated to continue treatment that isn't helping.
Who this fits best
Ketamine works best for patients with: treatment-resistant depression (multiple SSRI failures), severe baseline symptoms, prominent hopelessness or suicidal ideation, no active substance use, motivation to engage with integration work, stable enough life circumstances to support treatment (housing, basic safety, some social support). Ketamine works less well for: severe ongoing trauma without trauma-specific therapy support, active untreated bipolar mania, active psychosis, severe ongoing substance use, or situations where the depression is primarily driven by ongoing untreated psychosocial stressors that overwhelm any medication. For these patients, ketamine may still help but as one component of a broader plan, not as the primary intervention.
Ready to start?
Tovani offers board-certified telehealth ketamine therapy for treatment-resistant depression, anxiety, PTSD, and chronic pain. Available in Florida and New Jersey.
Check eligibility5-minute screening · Reviewed by a board-certified physician · FL & NJ
Frequently asked
What's the response rate for ketamine?
Roughly 50-70% of treatment-resistant patients respond per clinical literature. The Murrough 2013 RCT documented 64% response within 24 hours of one dose. Real-world response rates with full induction protocols tend to fall in the upper half of this range when adequate integration support is included.
When should I conclude ketamine isn't working?
Not before 4-6 sessions; ideally not before 8-10 with at least one round of protocol adjustment (dose, frequency, or modality). Premature conclusions miss responders who needed adjustment. Conclusions after 8-10 well-conducted sessions are more reliable.
What predicts good response?
Treatment-resistance (multiple SSRI failures actually predict ketamine response well), more severe baseline depression, no active substance use, motivation to engage with integration, stable life context. Demographics matter less than these clinical and life-context factors.
If ketamine doesn't work, what's next?
TMS (especially for treatment-resistant depression), ECT (for severe cases), Spravato esketamine (sometimes converts non-responders to sublingual), formal trauma-focused therapy if trauma is the dominant driver, or combination approaches. Non-response doesn't close the door on treatment.
Should I stop other medications when starting ketamine?
Not necessarily. SSRIs and ketamine are compatible. Lamotrigine and some other medications may modulate response; your physician reviews your full medication list. Don't stop or change medications without prescriber guidance.
References
- Murrough JW et al. 2013, American Journal of Psychiatry. Ketamine RCT in treatment-resistant depression — 64% response within 24 hours vs 28% placebo, establishing the ~60-70% response rate benchmark in the responder population. PMID 23982301
- Sanacora G et al. 2017, JAMA Psychiatry. APA consensus on ketamine in mood disorders — discusses response prediction, treatment failure handling, and decision points for adjusting or switching protocols. PMID 28249076
- Janik A et al. 2025, JAMA Psychiatry. Esketamine monotherapy RCT in treatment-resistant depression — supports rapid antidepressant response in patients failing multiple prior treatments, relevant for predicting who responds to NMDA-glutamate modulation. PMID 40601310
- Mathai DS et al. 2024 — at-home telehealth ketamine treatment. Longitudinal study of at-home ketamine therapy for depression — characterizes real-world response patterns, including predictors and trajectories in supervised telehealth settings. PMID 38810787