Health calculators

Gail Risk Calculator

Updated Sep 4, 2026 By Infinity Calculator
Rate Formulas Sources Medical disclaimer
Eligibility Screening
Does this patient have a personal history of breast cancer, ductal carcinoma in situ (DCIS), lobular carcinoma in situ (LCIS), or chest radiation for Hodgkin lymphoma?
Answer "Yes" only for a confirmed diagnosis or prior chest radiation treatment.
Does this patient have a known BRCA1 or BRCA2 mutation, or a genetic syndrome linked to high breast cancer risk (Li-Fraumeni, Cowden, Bannayan-Riley-Ruvalcaba)?
Select "Unknown" if genetic testing has not been done.
Demographics
Valid range: 35–85 years. The Gail Model has only been validated for women aged 35 to 85.
years
Used only when "Asian American" is chosen below. Sub-group changes the baseline incidence rates.
Race / ethnicity
Race-specific incidence rates and risk coefficients are applied to the estimate.
Patient & Family History
At what age did this patient have her first menstrual period?
If unknown, select "Unknown" — this field will not penalize the estimate.
How old was this patient at her first live birth?
Include only live births. Select "No births" if this patient has never had a live birth (nulliparous). If unknown, "Unknown" defaults to the 25–29 reference category as a neutral placeholder for the estimate.
How many first-degree relatives has this patient had with breast cancer?
Include only mother, sisters, and daughters. Do not include aunts, grandmothers, or cousins. If unknown, "Unknown" defaults to 0 first-degree relatives, as this is the most common value in the population.
Has this patient ever had a breast biopsy with a benign (non-cancerous) result?
Benign biopsies only — cancerous results are handled by the eligibility screen above.
How many previous benign breast biopsies has this patient had?
Used in the estimate only when the biopsy answer above is "Yes".
Was atypical hyperplasia (atypical ductal or atypical lobular hyperplasia) found in any biopsy?
Used in the estimate only when the biopsy answer above is "Yes".
5-Year Risk of Invasive Breast Cancer
Patient's 5-year risk
—
 
Average risk for women her age & race/ethnicity
—
 
Lifetime Risk (to Age 90)
Patient's lifetime risk to age 90
—
 
Average lifetime risk for women her age & race/ethnicity
—
 
Step-by-Step Solution
Inputs Used
Summary of Inputs Used in This Calculation
Question Answer Provided

Introduction

The Gail Risk Calculator estimates a woman's chance of getting invasive breast cancer. It gives two numbers: her risk over the next 5 years, and her risk up to age 90. It also shows the average risk for other women her same age and race, so you can compare.

The tool uses the Gail Model, also called the Breast Cancer Risk Assessment Tool. It asks a few simple questions: current age, race or ethnicity, age at first period, age at first live birth, how many mothers, sisters, or daughters had breast cancer, and any past benign breast biopsies. Then it does the math and shows each step.

Doctors often use a 5-year risk threshold as the point to talk about risk-lowering medicine: the Breast Cancer Prevention Trial enrolled women aged 35 to 59 whose 5-year predicted risk was at least 1.66%.5 This calculator flags 1.70%, its own rounding of that threshold.

The Gail Model does not fit everyone. It is built for women ages 35 to 85. It should not be used for women with a past breast cancer, DCIS, LCIS, chest radiation, or a known BRCA1 or BRCA2 gene change. Those women need a different model, such as the Tyrer-Cuzick Calculator. This tool checks for that first and will tell you if the Gail Model is not the right choice.

This calculator is a guide, not a diagnosis. Use it with a doctor to help plan screening and next steps.

How to use our Gail Risk Calculator

Enter a few details about the patient's age, race, health history, and family history of breast cancer. The Gail Risk Calculator then shows her 5-year risk and lifetime risk (to age 90) of invasive breast cancer, compares both to the average woman her age, and shows the math step by step.

Personal history of breast cancer, DCIS, LCIS, or chest radiation: Pick "Yes" only if she has one of these confirmed. If you pick "Yes," the tool stops, because the Gail Model does not work for these patients.

BRCA1 or BRCA2 mutation or genetic syndrome: Pick "Yes," "No," or "Unknown." Pick "Unknown" if she has never had genetic testing. "Yes" stops the tool, since these patients need a hereditary risk model instead.

Current age: Type her age in whole years. The number must be between 35 and 85, because the Gail Model was only tested in that age range.

Asian American sub-group: Choose her country or region of origin. This only counts if you pick "Asian American" for race, and it changes the baseline cancer rates used.

Race / ethnicity: Pick the one that fits her best. Each choice uses its own breast cancer rates and risk numbers, so this changes the result.

Age at first menstrual period: Pick the age range when her periods started. Choose "Unknown" if you do not know; this will not raise her score.

Age at first live birth: Pick the age range for her first live birth. Choose "No births" if she has never given birth, or "Unknown" if you do not know.

First-degree relatives with breast cancer: Count only her mother, sisters, and daughters. Do not count aunts, grandmothers, or cousins.

Previous benign breast biopsy: Pick "Yes" if she has ever had a breast biopsy that came back non-cancerous. Pick "No" or "Unknown" if not.

Number of benign biopsies: Pick 1 or 2 or more. This is only used if you answered "Yes" to the biopsy question.

Atypical hyperplasia: Pick "Yes" if a biopsy showed atypical ductal or lobular hyperplasia. This is also only used when the biopsy answer is "Yes."

Calculate and Reset: Results update as you answer, but you can click Calculate at any time. Click Reset to clear everything and start a new patient.

What Is the Gail Model?

The Gail Model is a tool doctors use to guess a woman's chance of getting invasive breast cancer. It was built by Dr. Mitchell Gail and other researchers at the National Cancer Institute, who published it in 1989 from case-control data of the Breast Cancer Detection Demonstration Project.2 It is also called the Breast Cancer Risk Assessment Tool (BCRAT), and the NCI's version draws on data from 280,000 white women aged 35 to 74 in that project plus SEER incidence data.4

The model gives two numbers: the chance of breast cancer in the next 5 years, and the chance by age 90 (lifetime risk). It compares those numbers to the average woman of the same age and race.

What the Model Looks At

The Gail Model uses a small set of facts about a woman's health and family:

  • Age. Risk goes up as women get older.
  • Race or ethnicity. Breast cancer rates are not the same in every group.
  • Age at first period. Starting young means more years of hormone exposure.
  • Age at first live birth. Having a first baby later, or not at all, raises risk a little.
  • First-degree relatives with breast cancer. Only a mother, sister, or daughter counts.
  • Past breast biopsies that were not cancer. The number of biopsies matters.
  • Atypical hyperplasia. These are odd-looking but non-cancer cells found in a biopsy. They raise risk.

The 1.7% Threshold

A 5-year risk of 1.7% or higher is the line many guidelines use. It comes from the Breast Cancer Prevention Trial, which enrolled women whose 5-year predicted risk was at least 1.66%.5 Women at or above that line are often high enough risk to talk with their doctor about risk-lowering medicine, like tamoxifen or raloxifene. This is called chemoprevention. Being above the line does not mean a woman will get cancer. It means a talk about extra screening or prevention makes sense.

Who the Gail Model Is Not For

The model does not work for everyone. It should not be used for women who:

  • Already had breast cancer, DCIS, or LCIS
  • Had chest radiation for Hodgkin lymphoma
  • Carry a BRCA1 or BRCA2 gene change
  • Have a syndrome like Li-Fraumeni, Cowden, or Bannayan-Riley-Ruvalcaba
  • Are under 35 or over 85

These women need a different model, such as Tyrer-Cuzick, BOADICEA, BRCAPRO, or Claus, and often need to see a genetics specialist.

What the Model Misses

The Gail Model is simple, so it leaves things out. It does not use breast density, weight, alcohol use, hormone therapy, or family history on the father's side. It also does not count aunts, grandmothers, or cousins. Research in Black women led to a separate model (the CARE model), which usually gave higher risk estimates than the original tool and is now recommended for counseling Black women.1 Data is limited for American Indian and Alaska Native women.

How to Use the Result

The invasive-cancer version of the model was validated in 5,969 women in the Breast Cancer Prevention Trial's placebo arm, where the ratio of expected to observed cancers was 1.03.3 The score is a group average, not a promise about one person. Two women with the same score can have very different outcomes. Use the number as a starting point for a talk with a doctor about mammogram timing, MRI screening, genetic testing, or prevention medicine. Never start or stop treatment based on this number alone.

Formulas used

Linear predictor (age < 50)
LP_1 = \beta_1 N_{biopsy} + \beta_2 A_{menarche} + \beta_3 A_{first\,birth} + \beta_4 N_{relatives} + \beta_5 A_{first\,birth} N_{relatives} + \ln(RR_{hyperplasia})
Linear predictor (age ≥ 50)
LP_2 = LP_1 + \beta_6 N_{biopsy}
Relative risk
RR = e^{LP_1} \;\; (age \lt 50), \qquad RR = e^{LP_2} \;\; (age \ge 50)
Patient-specific breast cancer hazard (attributable-risk adjusted)
h_1(a) = \lambda_1(a) \times (1-AR) \times RR, \qquad h_2(a) = \lambda_2(a)
Absolute risk over the interval from age t_1 to t_2
Risk = \sum_{a=t_1}^{t_2-1} \frac{h_1(a)}{h_1(a)+h_2(a)}\left(1-e^{-\left(h_1(a)+h_2(a)\right)}\right) e^{-\sum_{j=t_1}^{a-1}\left(h_1(j)+h_2(j)\right)}
Average (population) risk: same summation with RR = 1 and (1 - AR) = 1
Risk_{avg} = \sum_{a=t_1}^{t_2-1} \frac{\lambda_1(a)}{\lambda_1(a)+\lambda_2(a)}\left(1-e^{-\left(\lambda_1(a)+\lambda_2(a)\right)}\right) e^{-\sum_{j=t_1}^{a-1}\left(\lambda_1(j)+\lambda_2(j)\right)}
Risk ratio versus average, and 5-year / lifetime endpoints
Ratio = \frac{Risk_{patient}}{Risk_{avg}}, \qquad t_2 = \min(age+5,\,90) \;\text{or}\; 90
Chemoprevention threshold comparison
Risk_{5} \ge 1.70\%

Frequently asked questions

What is a first-degree relative?

A first-degree relative is a mother, sister, or daughter. These share about half of their genes with the patient.

Aunts, grandmothers, cousins, and half-sisters do not count in the Gail Model. Neither does the father's side of the family. Only count blood relatives with breast cancer, not other cancers.

What happens if I pick Unknown for a question?

Each Unknown answer uses a safe default so the math still works:

  • First period: treated as age 14 or older, the lowest-risk group
  • First live birth: treated as ages 25 to 29
  • Relatives with breast cancer: treated as 0
  • Biopsy: treated as no biopsy
  • Atypical hyperplasia: treated as neutral, a factor of 1.00

Unknown answers never raise the score. Real answers give a better estimate.

Why does No births give the same value as ages 25 to 29?

In the Gail Model, never giving birth carries about the same risk as a first birth in the late 20s. So the tool maps both to the same category. This matches how the original National Cancer Institute tool handles it.

Why does lifetime risk stop at age 90?

The Gail Model's incidence tables end at age 90. Very few new cases are counted past that point, and death from other causes is already built into the math. So the tool adds up yearly risk from the patient's current age through age 89.

What does the number like 1.45× the average mean?

It compares the patient to other women of the same age and race with no risk factors. A 1.45× result means her risk is about 45% higher than that average.

This is a ratio, not a percentage. A high ratio on a small base is still a small chance. Always read the percent number next to it.

Why does changing race or ethnicity change the result?

Breast cancer rates are not the same in every group. The tool swaps in different baseline rates and different risk weights for White, Black, Hispanic, and Asian American women. The NCI's estimates for Black women come from the CARE study of 1,607 women with invasive breast cancer and 1,647 without, plus SEER data.4

American Indian, Alaska Native, and Unknown answers use White rates, because there is not enough separate data. Those results are less exact.

What is atypical hyperplasia and why does it matter so much?

Atypical hyperplasia means a biopsy found odd-looking cells that are not cancer. There are two kinds: atypical ductal hyperplasia (ADH) and atypical lobular hyperplasia (ALH).

In this tool, a Yes answer multiplies risk by 1.82. A No answer lowers it slightly to 0.93. Unknown leaves it at 1.00.

Does the biopsy answer only count if I say Yes?

Yes. The number of biopsies and the atypical hyperplasia answer are only used when the biopsy question is set to Yes. If you pick No or Unknown, the tool ignores both and shows a note telling you so.

Is a high score the same as having breast cancer?

No. The score is a chance, not a diagnosis. Most women above the 1.7% line never get breast cancer, and some women below it do.

Only a doctor, imaging, and a biopsy can find cancer. Use the score to plan screening and talk about prevention.

What is invasive breast cancer?

Invasive breast cancer has spread out of the milk ducts or lobules into nearby breast tissue. This calculator only estimates invasive cancer.

It does not estimate DCIS, which stays inside the duct, and it does not estimate death from breast cancer.

Should I use lifetime risk to decide about a breast MRI?

No. Many guidelines use a 20% lifetime risk to offer MRI, but they say not to use the Gail Model for that call. Gail leaves out too much family history.

Use a model like Tyrer-Cuzick, BOADICEA, or Claus for MRI decisions, and ask a doctor.

How often should I run the calculation again?

About once a year, or any time something changes. Age alone raises the score every year. Redo it if a mother, sister, or daughter is diagnosed, or after a new benign biopsy.

Why does the Asian American sub-group matter?

Breast cancer rates differ a lot between Chinese, Japanese, Filipino, Hawaiian, and Pacific Islander women. Picking the right sub-group loads the right baseline rates.

If you pick Unknown, the tool uses Other Asian rates. The sub-group is ignored unless race is set to Asian American.

Can I use this after a mastectomy or breast implants?

No for a mastectomy done to treat cancer, since that falls under personal history and the tool will stop. Risk after a preventive mastectomy is also not covered by the Gail Model.

Implants alone do not change the score, but they can make mammograms harder to read. Tell the imaging team.

Why is my risk only a little above average even with a family history?

One relative raises risk, but age drives most of the number. A younger woman starts from a very low base, so the total stays small even after the increase. Look at both the percent and the ratio to see the full picture.

Sources

  1. Gail MH, Costantino JP, Pee D, Bondy M, Newman L, Selvan M, et al. Projecting individualized absolute invasive breast cancer risk in African American women. Journal of the National Cancer Institute. 2007;99(23):1782-1792. doi:10.1093/jnci/djm223. Accessed September 1, 2026.
  2. Gail MH, Brinton LA, Byar DP, Corle DK, Green SB, Schairer C, Mulvihill JJ. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. Journal of the National Cancer Institute. 1989;81(24):1879-1886. doi:10.1093/jnci/81.24.1879. Accessed September 4, 2026.
  3. Costantino JP, Gail MH, Pee D, Anderson S, Redmond CK, Benichou J, Wieand HS. Validation studies for models projecting the risk of invasive and total breast cancer incidence. Journal of the National Cancer Institute. 1999;91(18):1541-1548. doi:10.1093/jnci/91.18.1541. Accessed September 4, 2026.
  4. About the Breast Cancer Risk Assessment Calculator (The Gail Model). National Cancer Institute. Accessed September 4, 2026.
  5. Fisher B, Costantino JP, Wickerham DL, Redmond CK, Kavanah M, Cronin WM, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. Journal of the National Cancer Institute. 1998;90(18):1371-1388. doi:10.1093/jnci/90.18.1371. Accessed September 4, 2026.

Related Health calculators

All Health calculators
Health

Breast Cancer Risk Calculator

Estimate 5-year and lifetime breast cancer risk using the Gail model with Tyrer-Cuzick adjustments. Enter age, family history, density, and biopsy results.

Open Health

Tyrer-Cuzick Calculator

Estimate 10-year and lifetime breast cancer risk from BRCA, family history, density, BMI, and HRT. Shows risk category, population comparison, and charts.

Open