Facts, studies and product details checked in November 2026.
A typical adult woman's voice is about two-thirds higher in pitch than a typical adult man's. In a classic study of American English speakers, most of them from Michigan, women averaged about 220 vibrations a second (220 Hz) and men about 131 Hz. That is a little less than an octave, not a full one. Two things make the difference: longer vocal folds give men a lower pitch, and a longer throat and mouth give their voice a deeper resonance. We hear a voice as male or female from both together. But the two groups overlap, and before puberty boys and girls sound almost the same. For a lesson or a course, the research on narrator gender points both ways, and the later studies find no effect on learning. So choose the voice for the role and for how it reads your text, not for its gender.
We make one of the tools in this guide. Kukarella's Voice Library lets you filter 2,000+ voices by Gender and Age, and I marked almost every voice myself, by listening to it. But a filter is only a start: the sample you hear is a fixed recording, not your text, and hearing your own words in two voices uses credits twice. Each section below says what holds for any tool, and then what to do in Kukarella.
In this guide:
- What is the difference between a male and a female voice?
- Why are men's voices deeper?
- What makes a voice sound masculine or feminine? Hear it.
- Can a woman voice a man, or a man a woman?
- Do babies prefer female voices?
- Are female voices trusted more? Are male voices more authoritative?
- Male or female narrator: what the learning studies show.
- How to choose the narrator's voice for a lesson.
- Male and female voices in Kukarella.
- Frequently asked questions.
What is the difference between a male and a female voice?
The main difference is pitch: how many times a second the vocal folds vibrate. In 1995, James Hillenbrand and his colleagues at Western Michigan University recorded 45 men, 48 women and 46 children aged 10 to 12 reading the same words. Their data are public. The averages are below.
| Speakers | Average pitch | Range of speakers' averages |
| Men | 131 Hz | 96 to 192 Hz |
| Women | 220 Hz | 162 to 259 Hz |
| Boys and girls, 10–12 | 236 and 238 Hz | almost the same |
Three things in this table matter more than the averages.
- The gap is less than an octave. An octave doubles the pitch. Women's 220 Hz is 1.68 times men's 131 Hz, which is about three-quarters of an octave.
- The ranges overlap, at the edges. In this study the highest man averaged 192 Hz and the lowest woman 162 Hz: a few men speak higher than a few women.
- Children do not differ. At 10 to 12, boys and girls speak at nearly the same pitch. The difference arrives at puberty.
Average speaking pitch: men 96 to 192 Hz, women 162 to 259 Hz, children aged 10 to 12 about 237 Hz.These are read words, and people often speak a little lower in conversation. But the order holds: men lowest, then women, then children.
Why are men's voices deeper?
Because a man's vocal folds are longer, and his vocal tract is longer too. Ingo Titze, a voice scientist, compared male and female voices in 1989. He proposed that pitch scales mainly with the length of the vibrating part of the vocal folds, by a factor of 1.6. A longer string plays a lower note, and the vocal folds work the same way.
The second part is the "tube" above the folds: the throat, mouth and lips, called the vocal tract. A longer tube resonates lower, as a cello sounds deeper than a violin. MRI scans of 129 people aged 2 to 25 showed that the male and female vocal tracts do not differ in children and grow apart at puberty (Fitch and Giedd, 1999).
A shorter vocal tract and a longer one: the second part of why men's voices sound deeper.The change in boys comes late. A study that followed 26 boys through puberty found the voice "breaking" late in puberty (Harries and colleagues, 1997). Humans have the biggest difference in pitch between the sexes of any ape (Puts and colleagues, 2016). Why it evolved is still argued.
What makes a voice sound masculine or feminine?
A "masculine voice" is one that listeners hear as a man's. It is low in pitch and deep in resonance. Pitch alone is not enough. In a 2009 experiment, Hillenbrand and Michael Clark took recordings of men and women and shifted their pitch, their resonance, or both. Shifting both changed the speaker's apparent sex about 82% of the time. Shifting either one alone usually failed. So a high man still sounds like a man, and a low woman still sounds like a woman.
You can hear this with text to speech. The three recordings below read the same sentence, made in Kukarella with two US English voices, Jenny and Guy, from Microsoft. The first is the female voice and the second the male voice. In the third, the male voice's pitch is raised by the maximum our editor allows, 20%.
A female voice: Jenny, US English
The male voice reads the same sentence.
A male voice: Guy, US English
Now the same male voice, 20% higher.
The same male voice, pitch raised by 20%
We measured the three recordings with a simple pitch tracker. The female voice speaks at about 175 Hz, the male voice at about 160 Hz, and the raised male voice at about 195 Hz. So the raised male voice is higher than the female one, and it still sounds like a man: pitch has moved, resonance has not. To my ear it is a very strange man, speaking almost in falsetto. But a man. These two voices are also close in pitch to begin with, which is the overlap from the table above.
Gender is heard, and people do not always hear it the same way, even the same person. In autumn 2026 I listened to the voices in our library and marked how each one sounds. Some voices that the providers had filed as male sounded to me like women, and a few like young girls. We corrected the labels. One voice I heard in August as a teenage boy, I heard two months later as an adult I could not place as male or female. I had not looked at my first answer.
Can a woman voice a man, or a man a woman?
Yes, and audiobook narrators do it in every book with two sexes in it. They do it less than you might think. A study of dialogue passages from LibriVox audiobooks of Dickens, read by men and women in equal numbers, found that readers raised their pitch only slightly for female characters, and did not lower it for male ones (Stolarski, 2017). The narrator and author Mary Robinette Kowal warns that at a live reading, a cross-gender voice that is very bad, or very good, pulls the listener out of the story for a moment. Her advice is to colour every character's voice to the same degree, and to be subtle.
So a woman voicing a man changes her manner, rhythm and resonance a little, not only her pitch. With text to speech you do not need to: you choose a male voice for the man's lines. In a dialogue tool, that means a voice per speaker (see the Kukarella section below).
Do babies prefer female voices?
No. Babies prefer their mother's voice, which is a familiar voice, not a female voice in general.
- Newborns learned to suck on a dummy in a particular rhythm to hear their own mother's voice, and chose it more often than another woman's (DeCasper and Fifer, 1980).
- Newborns showed no preference for their father's voice over another man's (DeCasper and Prescott, 1984). The authors put the difference down to what babies hear before birth: the mother's voice.
- Before birth, a fetus's heart rate rose to its mother's voice and fell to a stranger's (Kisilevsky and colleagues, 2003, 60 fetuses).
Babies do prefer the sing-song way adults talk to babies. A study across 67 labs confirmed it in 2020 (ManyBabies Consortium), but the recordings were of mothers, so it says nothing about male or female voices.
Are female voices trusted more? Are male voices more authoritative?
Both claims are common, and neither is a fact about voices in general. The studies show stereotypes that people apply, and they point both ways.
- Toward male voices. In 2000, people applied gender stereotypes even to clearly synthetic voices, and found a male computer voice more persuasive (Lee, Nass and Brave; we cover this in How to Choose the Right AI Voice). In 2003, listeners rated male synthetic speech as preferable to female (Mullennix and colleagues).
- Toward female voices. In 2016, the CXL Institute played videos with four voice-overs to 202 people; they rated the video more trustworthy when a woman spoke the voice-over. In 2019, 240 people in the US rated one synthetic female voice 12.5% higher than one synthetic male voice (Voicebot). Its author called the result "by no means decisive".
- Toward lower pitch, in both sexes. When researchers raised men's voices by 20%, listeners judged them less truthful and more nervous (Apple and colleagues, 1979). Asked to choose a leader, men and women chose the lower-pitched version of both male and female voices (Klofstad and colleagues, 2012).
Each of the first two kinds of study used one or two voices per gender. That makes the result a fact about those voices, as much as about gender.
Advertising has long chosen men. In more than 2,000 ads shortlisted or awarded at the Cannes Lions festival from 2006 to 2016, 18% had only male voices and 3% only female (Geena Davis Institute and J. Walter Thompson, 2017). But a Spanish study of 372 journalism students concluded that this preference for male voices in radio ads "is more based on tradition than on alleged advertising effectiveness" (Rodero and colleagues, 2013).
A man on TV sells diapers, and the mother who buys them listens. Advertising has long chosen male voices out of habit.Voice assistants went the other way. A 2019 UNESCO report found Siri female by default in most markets, but male by default in British English, Arabic, Dutch and French. It concluded that studies "refute or complicate the idea that humans have a blanket preference for female voices".
Male or female narrator: what the learning studies show
For a teacher, this is the question that matters, and the answer is: it does not seem to matter much.
| Study | Learners | What they found |
| Linek and colleagues, 2010 | two experiments | Female narrator: better learning |
| García Rodicio, 2012 | 40 students | Male narrator: better learning |
| Hoogerheide and colleagues, 2016 | 167 secondary students | No effect of gender |
| Schrader and colleagues, 2021 | 95 students | No effect of gender |
The two earlier studies point in opposite directions. The two later ones, with 167 and 95 learners, found no difference. Schrader's team found that "speaker gender and speaker–learner gender similarity had no impact on learning gains". A 2022 review of 59 articles on on-screen teaching agents leaned towards male agents, then concluded that "prior research is not robust enough to prove the superiority of a male agent in all fields and for all audiences" (Armando and colleagues).
Most of these studies used one voice per gender. A pleasant female voice against a stiff male one tests the two voices, not the two genders. What research does show more often is that the quality of the voice matters: a human voice helped learners more than a machine-synthesised one (Mayer and colleagues, 2003). Today's best synthetic voices close much of that gap; we explain this in How to Make Text to Speech Sound Less Robotic.
So for a narrator, choose a voice that reads clearly and naturally. Then let the gender follow the role.
How to choose the narrator's voice for a lesson
Treat gender as casting, not as a learning trick. Five rules, in this order.
- Who is speaking in the text? If the narrator says "I" and is a woman, use a female voice. If the text is a man's story, use a male one. A gender that contradicts the text distracts.
- In a dialogue, make the speakers easy to tell apart. A man and a woman, or two speakers of different ages, are easier for learners to follow than two similar voices. A dialogue tool that casts voices for you will only make them differ if you ask: give each speaker a gender and an age.
- Think of the voices your learners will meet. A nurse in training will hear male and female patients. A learner of French will hear both. Give them both over a course.
- Keep one narrator for a whole course. Changing the narrator between modules sounds like a different course. Pick once, and keep it.
- Choose by ear, with your hardest line. Play the voice on your own text, not its sample. If you cannot decide, do what I have suggested since 2023: make both versions and let a few of your listeners hear them.
The questions above take a minute. The card at the end says which voice to start with, what to listen for, and how to find it.
When gender should not decide. A clear voice in the right accent matters more than its gender. A Mexican Spanish lesson needs a voice from Mexico more than it needs a woman or a man. And if your learners know you, your own voice may be the best narrator of all.
Male and female voices in Kukarella
In Kukarella, you find male and female voices in the Voice Library, which opens from Text to Voice and from DialoguesAI. Here is how, with the limits.
- Filter by Gender and Age. The filters are in the library's left column (behind a Filters button on a narrow screen). Under Gender, choose Female or Male. Under Age, choose one or more of Child, Teen, Adult and Senior. Each voice shows how it sounds, for example Female · Adult · 25–45.
- The labels come from a person, not from the provider. In October 2026 I listened to almost every voice in the library and marked its gender and age by ear. It was very tedious and very long. The filter uses my marks. But a few voices added since then carry no mark yet, and appear under no age.
- Press ▶ to hear a voice before you choose it, even a voice your plan does not open. But that is a fixed sample, not your text.
- To compare, make the same text in both voices. Each version uses credits for the whole text, so comparing two voices costs the text twice.
- Text to Voice reads the whole project in one voice. For a man and a woman in one script, use DialoguesAI: each speaker gets their own voice. Write the speakers' gender and age into your request, for example "a boy of nine and a woman of about forty". But the assistant casts from a smaller set of voices: if a speaker sounds wrong, open that speaker's card and choose another voice from the library.
- Pitch moves a voice a little, not to the other gender. You can raise or lower it by up to 20%, as in the recordings above. Not every voice takes it: tick Effects → Pitch in the filters to see only those that do.
- Child voices are mostly girls. Boys' voices are few. If a lesson needs a boy, filter Age → Child, Gender → Male, and play each one.
- "Neutral" means a voice I could not place as male or female. There are only a handful, and none is a native English voice. No voice in the library is offered as non-binary.
The Voice Library with the Gender filter open: All, Female, Male, Neutral, and each voice tagged with how it sounds.Try it in Text to Voice: open the library, set Gender, and play the samples.
Frequently asked questions
Who has a higher voice, males or females?
Adult women, on average. In the Hillenbrand study, women averaged 220 Hz and men 131 Hz. But the ranges overlap, and some men speak higher than some women.
Why do girls have higher voices than boys?
Before puberty they do not: at 10 to 12, boys and girls averaged 236 and 238 Hz. At puberty, boys' vocal folds and vocal tract grow longer, and their voices drop.
What frequency is a male voice? A female voice?
An adult man speaks at about 100 to 190 Hz, and an adult woman at about 160 to 260 Hz. Those are the ranges of speakers' averages in one well-known study. Each voice also moves up and down around its own average as it speaks.
Is a woman's voice an octave higher than a man's?
No, a little less: about three-quarters of an octave on average.
Can a man have a naturally female-sounding voice?
Yes. The ranges overlap, and a voice is heard as male or female from pitch and resonance together. A man with a higher pitch and a lighter resonance can be heard as a woman, and the reverse.
Is a male or a female voice better for e-learning?
Neither, on the evidence. The later studies found no effect of the narrator's gender on learning. The quality of the voice, its accent and its fit with the role matter more.
Why are most voice assistants female?
Mostly by choice of the companies and by habit. A 2019 UNESCO report found the evidence for a general preference for female voices weak, and noted that Siri used a male voice by default in some languages.