By Jordan Price · Growing Guide · Updated May 2026
- The science is inconclusive, not confirmatory: Conley et al. (2019) — the most rigorous published research on cannabis flushing — found no statistically significant difference in terpenes, cannabinoids, or consumer-rated taste between flushed and unflushed plants in a controlled hydroponic setting.
- Flushing logic is medium-specific: Coco coir accumulates calcium and magnesium displacement salts due to cation exchange — a real, measurable problem that flushing addresses. Organic soil has no synthetic salts to remove and flushing actively damages beneficial microbiology.
- TDS runoff measurement is the objective indicator: If runoff TDS is 500+ ppm above input water, salt accumulation is real and flushing has a logical basis. Without this measurement, flushing is speculative habit rather than evidence-based practice.
- Organic grows should never be flushed: Flooding a living soil with plain water leaches soluble nutrients and kills beneficial microorganisms — the opposite of what organic growers want in the final weeks. The microbial community is what makes organic grows taste exceptional, not flushing.
- Hydro systems flush faster than soil: In deep water culture, switching to plain water for 3–7 days effectively clears the root zone of residual salts because the roots are always in direct water contact. Soil and coco require 1–2 weeks to achieve the same runoff TDS reduction.
- Yellowing during flush is nitrogen deficiency, not nutrient purging: The widespread belief that leaf yellowing during a flush indicates the plant is burning stored nutrients for a cleaner smoke is not supported by research. Yellow leaves = nitrogen starvation. Cannabis does not store nitrogen in tissue the way tobacco stores chlorophyll-bound nitrogen.
- Trichome assessment remains the gold standard for harvest timing: Whether you flush or not, harvest based on trichome maturity (70–90% cloudy, some amber) — not on a calendar count from the flush start date.
What Is Cannabis Flushing?
Cannabis flushing is the practice of running large volumes of plain, nutrient-free water through a growing medium in the final 1–2 weeks before harvest. The theory behind it is that accumulated mineral salts from weeks of synthetic nutrient feeding are stored in both the growing medium and in plant tissue, and that removing these salts produces a smoother, cleaner-tasting, or more aromatic final product. The practice is widely recommended in cannabis cultivation guides and has been the dominant conventional wisdom in cannabis cultivation for decades.
The problem is that when cannabis flushing was finally subjected to controlled scientific testing, the results were significantly less clear-cut than the industry assumed.
Understanding the flushing debate requires separating two distinct claims: (1) Does flushing remove salts from the growing medium? Yes — this is measurable with a TDS meter and clearly true. (2) Does removing those salts improve the taste, smoothness, or terpene content of the final product? This is where the research fails to provide a clear answer, and where grower experience diverges strongly along medium and nutrient-regime lines.
The Research: What Studies Actually Show
The most frequently cited scientific examination of cannabis flushing was published by J. Conley and colleagues in 2019. The study grew cannabis in controlled hydroponic conditions, subjected one group to a standard two-week plain-water flush before harvest, left a second group on full nutrient feeding until harvest, and compared the two groups across multiple quality metrics: terpene concentration, total cannabinoid content, chlorophyll levels, and sensory evaluation panels of experienced cannabis consumers.
The key finding: no statistically significant difference was detected between flushed and unflushed plants across any of the primary quality metrics. Consumers in the blind taste panel could not reliably distinguish flushed from unflushed cannabis. Terpene profiles were equivalent. THC and CBD concentrations were equivalent. The study was widely reported as definitively disproving the value of flushing.
The research has meaningful limitations, however. It was conducted in hydroponic systems, which have fundamentally different salt-accumulation dynamics than soil and especially coco coir. It used a specific nutrient regime and may not reflect the outcomes of heavy-salt commercial grows. Sensory evaluation of cannabis taste is notoriously difficult to blind and control. The honest scientific conclusion is that flushing’s benefit in improving flavor remains unproven, not definitively disproven for all media and conditions.
| Evidence Type | Method | Finding | Limitation / Caveats |
|---|---|---|---|
| Conley et al. 2019 | Hydroponic RCT, consumer panel | No significant difference flushed vs. unflushed in terpenes, cannabinoids, or taste | Hydro only; specific nutrient line; small consumer panel; blinding difficult |
| TDS runoff measurement | EC/TDS meter on medium runoff | Measurably reduces salt concentration in coco and soil media — this effect is real and consistent | Does not prove reduced medium salts = better taste in plant tissue |
| Grower blind taste panels | Consumer sensory evaluation (anecdotal) | Mixed: some report improvement, many find no difference, results vary by medium | Not controlled, not randomised; strong confirmation bias; placebo effect likely |
| Organic grow comparisons | Observational reports from no-till community | Water-only and organic grows consistently report excellent taste without flushing | No comparable synthetic-nutrient control group; may reflect medium and practice differences |
| Substrate chemistry studies | Analytical chemistry of plant tissue vs. medium | Limited cannabis-specific data; general horticultural research suggests most mineral salts are in medium, not plant tissue | Cannabis-specific data lacking; extrapolation from other crops |
The Cannabis-is-Not-Tobacco Argument
The original rationale for flushing cannabis borrowed from tobacco cultivation, where flushing before harvest is a well-documented practice that reduces chlorophyll-bound nitrogen in leaf tissue and produces a smoother, more aromatic smoke. This mechanism is real in tobacco because tobacco stores significant amounts of nitrogen in chlorophyll bound to leaf tissue, and that nitrogen contributes to harshness when burned.
Cannabis physiology is different. Cannabis does not store nutrients in tissue the same way. The minerals present in cannabis flowers at harvest — primarily potassium, calcium, phosphorus, and residual nitrogen — are present largely as a result of what was in the growing medium at the time of harvest, not as stores that accumulated over weeks and can be drawn down by the plant. When you stop feeding and run plain water, the plant experiences nutrient deficiency — yellowing and leaf death — but this is a sign of deprivation, not a sign that stored nutrients are being mobilized and purged in a beneficial way.
The post-flush yellowing that many growers point to as evidence that flushing is working is simply nitrogen starvation. The nitrogen that may contribute to harsh smoke is in the medium as residual salts — which flushing does remove from the medium. Whether that matters for taste is the unresolved question.
When Flushing Does Have Logical Justification
| Situation | Flushing Justified? | Reason |
|---|---|---|
| Confirmed salt buildup in coco coir (runoff TDS 500+ ppm above input) | Yes | Cation exchange actively accumulates Ca and Na in coco; flushing with clean water replaces bound ions; measurable and logical |
| Synthetic nutrient heavy-feed regime in soil (runoff TDS chronically elevated) | Yes (optional) | Salt accumulation measurable; flushing before harvest reduces medium salt levels regardless of whether it affects taste |
| pH correction flush (runoff pH outside 5.8–7.0 range) | Yes | Flushing with correctly pH’d water resets medium pH; this is a maintenance flush, not a pre-harvest flush |
| Hydroponic DWC/NFT with synthetic nutrients at harvest | Optional | Switching to clean water for 3–7 days is low-risk and theoretically removes residual salts from root zone rapidly |
| Organic living soil / no-till / Korean natural farming | No — counterproductive | No synthetic salts present; flushing destroys microbial community and leaches organic nutrient reservoir; damages the product |
| Autoflowering strains near harvest | Only if runoff TDS elevated | Compressed lifecycle means extended flush causes visible nutrient deficiency and potential yield loss; maximum 5–7 days |
Medium-Specific Flushing Protocol
| Medium | Salt Risk | Flush Recommended? | Duration | Volume per Flush Session | Key Notes |
|---|---|---|---|---|---|
| Coco coir | HIGH | Yes, strongly | 1–2 weeks plain water | 2× container volume per session | Coco’s CEC binds Ca/Na; monitor runoff TDS to track progress; run until within 200 ppm of input |
| Synthetic soil (peat-based, heavy feeds) | MEDIUM–HIGH | Yes, optional | 7–10 days | 3× container volume per session | Light organic feeders on quality soil may not need flushing; check runoff TDS first |
| Deep Water Culture (DWC) | LOW–MEDIUM | Optional | 3–7 days | Drain reservoir; refill with clean pH’d water | Fast effect — roots in direct water contact; switch to R.O. or filtered water |
| NFT / Aeroponics | LOW–MEDIUM | Optional | 3–7 days | Full reservoir/line flush with clean water | Similar to DWC; clean lines and reservoir before introducing flush water |
| Organic living soil / no-till | NONE | No — counterproductive | N/A | N/A | Dolomite lime and organic amendments buffer nutrients; microbial cycling = excellent taste without flushing |
| Rockwool (commercial hydro) | MEDIUM | Yes | 5–7 days | Full nutrient solution replacement with clean water | Keep rockwool pH 5.5–6.0 during flush; salt can accumulate between cube layers |
TDS/EC Runoff Monitoring Protocol
The objective approach to flushing decisions is to measure, not assume. A TDS or EC meter used on runoff water tells you exactly what is happening in your medium.
- Measure input water TDS: Record the TDS of your water before any nutrients are added. This is your baseline. Clean tap water typically runs 50–200 ppm; R.O. water runs 0–20 ppm.
- Collect runoff sample: After watering normally, catch the first 50–100ml of water that drains from the pot. This represents the concentration in the medium root zone.
- Compare runoff TDS to input TDS: If runoff TDS is 200–400 ppm above input: normal, no immediate concern. If runoff is 500–800 ppm above input: salt accumulation present; flushing justified. If runoff is 800+ ppm above input: significant buildup; flush immediately.
- Flush until runoff converges with input: During a pre-harvest flush, run plain pH’d water through the medium until runoff TDS drops within 200–300 ppm of input water. This may take multiple sessions over 1–2 weeks for coco and soil.
- Do not harvest based on calendar — harvest based on trichomes: Continue trichome assessment throughout the flush period. Harvest when 70–90% of calyx trichomes are cloudy with some amber, regardless of how many days into the flush you are.
CO2 Flush vs. Water Flush: The Fringe Debate
Some cultivators argue that the purpose of pre-harvest flushing is not to remove medium salts at all, but to force a final CO2 depletion event where the plant consumes all available carbon substrates through respiration, producing a “cleaner” burning flower. This theory has essentially no scientific support and confuses plant physiology with medium chemistry. CO2 levels in a properly ventilated grow are determined by the environment, not by what nutrients the plant is fed. The CO2 flush theory is not considered valid by plant scientists.
Common Flushing Mistakes
| Mistake | Effect | Correct Practice |
|---|---|---|
| Flushing organic soil | Destroys beneficial microbes; leaches organic nutrient reservoir; can cause pH swings | Never flush organic growing media; trust the microbiology |
| Starting flush based on calendar, not trichomes | May begin flush too early (plant still building yield) or too late (already past peak) | Check trichomes daily from week 6 of flower; begin flush protocol only when trichomes are entering the maturity window |
| Not adjusting pH of flush water | Incorrect pH prevents effective salt dissolution and can cause pH lockout in medium | pH flush water to 6.0–6.5 (soil) or 5.8–6.2 (coco/hydro) |
| Flushing autoflowers for 2+ weeks | Compressed lifecycle means extended flush causes severe nutrient deficiency before harvest-ready | Maximum 5–7 days for autoflowers; only if runoff TDS confirms salt buildup |
| Interpreting yellowing as success | Grower reinforces incorrect flushing theory; continues harmful practice | Understand that yellowing = nitrogen starvation; it is not evidence that flushing is working |
| Skipping TDS measurement and flushing by default | May be flushing a medium with no salt buildup, causing unnecessary nutrient deficiency near harvest | Always measure runoff TDS before deciding to flush; evidence-based decision only |
FAQ: Cannabis Flushing
If flushing doesn’t improve taste, what does?
Taste quality in cannabis is primarily determined by: (1) genetics and terpene profile — the strain itself; (2) terpene preservation during harvest, drying, and curing (see our terpene preservation guide); (3) proper curing at correct humidity for adequate time; (4) harvest timing at peak trichome maturity; and (5) avoiding heat damage in any stage post-harvest. A slow-dry, long-cure cannabis grown without flushing will consistently outperform a fast-dried, poorly cured plant that was flushed for two weeks. If you’re tasting harshness, look to your dry and cure process before your nutrient program.
Should I flush coco coir even if my runoff TDS is acceptable?
If your runoff TDS is within 200–400 ppm of your input water throughout the grow, salt accumulation has been managed effectively through good fertigation practice (frequent enough waterings with some runoff each time). In this case, there is no measurable salt buildup to flush, and reducing nutrients for 1–2 weeks before harvest may cause unnecessary deficiency symptoms without addressing a real problem. Many coco growers on regular fertigation schedules find that simply reducing feed EC in the final 1–2 weeks (rather than eliminating nutrients entirely) achieves a natural taper without the nutrient starvation of a full flush.
Does flushing affect cannabinoid content?
The Conley 2019 study found no significant difference in THC or CBD levels between flushed and unflushed plants. This aligns with the plant physiology: cannabinoid biosynthesis continues in trichome heads until trichomes reach full maturity, and is driven by genetics, light intensity, trichome development timeline, and temperature — not by whether the plant received nutrients or plain water in the final 2 weeks. Extended flushing that causes severe deficiency stress in late flower can potentially reduce final yield as bud development continues during the flush period, but this is a result of nutrient deprivation affecting the plant’s growth, not a direct effect on cannabinoid concentration.