Voice depth is mostly genetic — vocal fold length and larynx position are set by puberty and you can't grow longer folds. But most untrained adults speak above their natural anatomical floor, and you can train your speaking pitch down by roughly 1–2 semitones with consistent work over a few weeks. That's small on paper and very noticeable in person.
Below: the anatomy that determines your ceiling, the technique that actually shifts your speaking pitch, what doesn't work (and what's dangerous), why some voices — including some women's — are naturally deeper, and how to measure where yours sits right now.
Three things, roughly in order of importance: vocal fold length, larynx position, and vocal tract length.
Vocal fold length sets your fundamental frequency (F0) — longer, thicker folds vibrate slower, lower pitch. Adult male folds are about 17–23 mm long; adult female folds about 12–17 mm. That difference is almost entirely a testosterone-driven puberty effect: male folds grow 4–11 mm (up to ~60% in length), female folds grow 1.5–4 mm. The research linking testosterone exposure to fold growth and lower F0 is solid — vocal fold length mediates most of the male/female pitch difference.
Larynx position matters almost as much. A lower-sitting larynx lengthens the vocal tract, which lowers your formants and gives the voice that "imposing" depth. Comparative anatomy work suggests the human male larynx descended evolutionarily partly to exaggerate apparent body size through deeper vocalizations.
Vocal tract length (lips to glottis) correlates with body height. A Journal of Voice study of laryngeal morphometry found positive correlations between body height and thyroid cartilage size, vocal fold length, and neck circumference at the Adam's apple. Taller people, on average, have slightly lower voices — but the effect is weaker than people think, because the human larynx is unusually free from skeletal constraints.
Typical adult speaking F0, per Hollien's reference data:
Anything well below your sex's typical range reads as "deep." A male speaking around 85 Hz reads as notably deep; a female speaking around 145 Hz reads the same way.
Your absolute floor — the lowest note you can sustain without vocal fry — is essentially anatomy. You don't grow your vocal folds longer after puberty. You don't lower your larynx by another centimeter in adulthood. The hardware is locked.
What you can change is where you speak inside that hardware. Most untrained adults sit above their natural floor by 2–4 semitones for habit reasons — they learned to project, they raised pitch under stress, they imitated someone, or they just never thought about it. Research on behavioral pitch lowering shows cisgender females can drop several semitones with structured therapy (VFE, resonant voice therapy, lip-rounding) over 4 weeks of daily practice. The same is true for men with habitually raised speaking pitch.
To calibrate expectations: puberty drops male F0 by about an octave (12 semitones) in 12–24 months. Training drops a habitually high speaking voice by 1–3 semitones over weeks to months. The first is anatomy; the second is coordination. The second is what's available to you.
This is the part most search results lie about. Here's the evidence-backed version.
The biggest lever. Open a piano app or tuner, find your current speaking pitch (read a paragraph, watch the Hz reading), then practice reading that same paragraph 1–2 semitones lower. Hold it for 5–10 minutes a day. After 4–6 weeks the new pitch becomes default.
This works because trained speakers use a wider portion of their phonational range than untrained speakers — most people are sitting well above their anatomical low without realizing it.
A lowered larynx lengthens the vocal tract and deepens resonance. The standard way to find the feeling: start a yawn and freeze mid-yawn — the larynx drops naturally. Sing or speak on "ng" or a hummed "mmm" from that position. The "tube breathing" approach used in transmasculine voice work uses the same mechanism.
Critical caveat: gently. Pulling the larynx down with throat tension just adds strain. The yawn cue is the right entry because it activates the laryngeal depressor muscles without engaging compensatory tension.
A collapsed forward-head posture compresses the throat, shortens vocal tract resonance, and forces a higher speaking pitch. Effects of body posture on voice range profile in healthy adults show measurable differences in voice production with postural alignment.
The fix is unglamorous: ribcage up, shoulders back, head neutral, breathe into the belly. Diaphragmatic breathing also pulls the larynx slightly downward as the trachea stretches, which deepens resonance.
Standardized voice therapy protocol — warm-up, sustained "ee," upward and downward glides, sustained pitches. Systematic reviews in the Journal of Voice report effect sizes from -0.59 to 1.55 across vocal performance measures, with moderate-to-strong improvement on voice self-report. VFEs don't grow your folds, but they improve the efficiency and coordination of the voice you have — which often translates to a steadier, lower-feeling speaking voice.
Dehydrated vocal folds need more pressure to vibrate (higher phonation threshold pressure, per a meta-analysis with average effect size 0.33). Underhydrated voices sound stiff, breathy, and slightly higher because of inefficient closure. Drinking water won't make your voice deeper than its baseline — but staying hydrated keeps you at that baseline instead of above it.
If you frequently default to a light, head-dominant speaking register, you're skipping the bottom half of your range. Practice reading aloud in clear modal/chest voice. Bass-heavy resonance lives there.
Realistic timeline: 15 minutes of daily practice for 4–8 weeks for a noticeable shift in speaking pitch. Three months for it to become unconscious. Not glamorous, but real.
Speaking below your natural floor causes vocal fold edema, fatigue, and over time, nodules and polyps — which paradoxically can make your voice sound less deep and more strained. If you feel throat tension, you're going too low.
It does deepen voices. Reinke's edema, a chronic vocal fold swelling, is found in 97% smokers among diagnosed patients. The fluid-filled folds vibrate slower and the voice drops. In severe female cases, the pitch lowers enough that callers mistake the speaker for a man.
It's also irreversible without surgery, comes packaged with cancer risk, and the deepening sounds gravelly and aged rather than authoritative. Trading your lungs for half an octave is a bad deal.
No clinical evidence for any of them. Saved you $40.
This works — meta-analyses of testosterone therapy for transmasculine voice masculinization show clear F0 lowering, though 21% don't reach cisgender male F0 ranges after a year and 24% need additional voice therapy. It's also a lifelong endocrine intervention with major systemic effects — only viable under medical supervision, not a self-improvement hack.
Sulfur hexafluoride is denser than air, so it lowers vocal resonance and produces an absurdly deep voice. It is also heavier than air, which means it pools in the lungs and doesn't easily exhale, displacing oxygen. The CDC NIOSH Pocket Guide lists it as a simple asphyxiant, and documented case reports include pulmonary edema and asphyxiation. People have died doing this on YouTube. Don't.
This is the "why is my voice so deep" / "why is my voice so deep as a girl" cluster. The honest answer covers several mechanisms, most of them not problems.
Voice depth is normally distributed within each sex. Some people land on the lower tail without anything being "wrong." Hollien's data shows healthy adult female F0 ranging from 165 to 294 Hz — that's roughly an octave of normal variation. A woman speaking at 170 Hz isn't pathological; she's at the low end of normal.
Taller people, on average, have slightly larger larynges and longer vocal folds. The correlation is real but weaker than online discourse suggests — body size accounts for some of the depth variance, not most of it. Genetics sets the rest.
Polycystic ovary syndrome involves elevated androgens, which can cause thyroarytenoid muscle hypertrophy and lower female F0. The Egyptian Journal of Otolaryngology PCOS analysis found significantly decreased mean F0 in PCOS patients versus controls, alongside vocal fatigue and frequent throat clearing. Effect sizes vary across studies — some find statistically significant lowering, others don't — but the mechanism is well-established. If you're a woman whose voice noticeably dropped post-puberty and you have other PCOS markers, that's worth raising with a doctor.
Other endocrine causes: congenital adrenal hyperplasia, androgen-producing tumors (rare), certain anabolic steroids and medications. Voice and endocrinology overviews cover the full list.
Trans men on testosterone experience clear F0 drops over months — same mechanism as natal puberty, just later.
All deepen the voice over time at any sex. Reinke's edema (smoking-induced) is the most dramatic example.
A few names that come up in this search — Brittney Griner, Lauren Betts, Miley Cyrus, Denise Richards, Lyn Alden — and an honest note about each.
Miley Cyrus has publicly confirmed she has Reinke's edema and a vocal cord polyp, explaining the condition in interviews. She had surgery in 2019 but has chosen not to fully resolve the polyp because the texture is part of her sound. That's the one case in this list with a public, on-the-record medical explanation.
Brittney Griner is 6'9", and the physiological explanation cited in coverage points to her larger vocal folds and thoracic cavity — natural anatomical variation consistent with her height. She has spoken publicly about being bullied for it.
For the others, we don't have on-the-record medical statements, so we shouldn't speculate. The range of explanations covered above — natural variation, height/build, hormonal factors, vocal use over years — covers the possibilities. Publicly diagnosing celebrities' larynges from a podcast clip is rude and usually wrong.
Practical, takes 60 seconds:
Reference benchmarks (Hollien's norms):
| Voice | Average F0 | "Deep" range |
|---|---|---|
| Adult male | 100–130 Hz | <100 Hz |
| Adult female | 165–220 Hz | <160 Hz |
Want to know exactly how deep your voice is? Run a recording through the Voice Depth Analyzer — it returns your average F0, a comparison to age/sex norms, and a depth percentile. Free, no signup, instant. For voice type (bass / baritone / tenor / alto / mezzo / soprano), pair it with the Voice Type Classifier. For tone, breath support, and pitch stability in the same upload, layer on Vocal Analysis. And if you're curious how your voice reads age-wise, the Voice Age Estimator covers that.
Run two or three samples at different times of day. Voices have a daily fatigue curve — morning F0 is often slightly lower than afternoon, and post-eight-hours-of-talking F0 shifts again. One measurement is a snapshot; three is a trend.
Quick aside since it comes up. Useful adjectives, ranked from descriptive to evocative: low, deep, resonant, bass-heavy, warm, rich, sonorous, baritone, gravelly, smoky, thunderous, booming. "Resonant" and "warm" describe vocal tract length and fullness; "gravelly" and "smoky" describe noise/jitter content; "booming" describes volume and projection more than pitch. For writing, pair a pitch word with a texture word ("deep and gravelly," "low and resonant") for specificity.
People chase deep voices for confidence, perceived authority, dating, professional reasons. None of those are stupid. But: deeper isn't objectively better. A confident, well-supported mid-range voice beats a strained, forced low voice every time — listeners hear the strain even when they can't name it.
The realistic target isn't "sound like Morgan Freeman." It's "speak at the bottom of your natural range, with good breath support, without strain." That alone reads as more grounded and authoritative than most untrained voices.
For the AI read on your own voice: the Voice Depth Analyzer returns your F0 with norms and percentiles, the Voice Type Classifier places you in a Fach category, Vocal Analysis covers tone and breath support, the Voice Age Estimator reads your perceived age, and the Voice Health Analyzer flags fatigue markers like elevated jitter and shimmer. All free, no signup.