Breast Cancer Risk Calculator
Answer the seven questions the National Cancer Institute’s Breast Cancer Risk Assessment Tool asks, and this calculator runs the published Gail model on your answers: your 5-year absolute risk of invasive breast cancer, your lifetime risk to age 90, and the average risk for a woman your age and race, side by side.
Read this first: this page implements the BCRAT (Gail model) mathematics exactly as published, with its race-specific baseline rates and coefficients. It is not the official NCI website, and no online result should drive a medical decision on its own. The model only applies to women aged 35 to 85 with no history of breast cancer, DCIS or LCIS, no known BRCA1 or BRCA2 mutation, and no chest radiation for Hodgkin lymphoma. If any of those apply, skip the calculator and talk to your doctor or a genetic counselor. Everyone else: use the result to arrive at that conversation informed, and cross-check it on the official NCI tool if you want a second rendering of the same model.
Population-based estimate from the published Gail model (BCRAT), not medical advice and not a diagnosis. With a strong family history of breast or ovarian cancer, talk to your doctor or a genetic counselor rather than relying on any online calculator.
Quick Answer
Key takeaways
- This is the real Gail model (BCRAT): published relative-risk coefficients, race-specific SEER baseline rates, competing mortality, and the attributable-risk adjustment, integrated year by year.
- Every result shows four numbers: your 5-year risk, the average 5-year risk for your age and race, your lifetime risk to 90, and the average lifetime risk.
- The 1.67% 5-year line is where clinical conversations about risk-reducing medication and enhanced screening usually begin. A lifetime risk of 20% or more is where annual MRI is commonly added to mammography.
- Hard eligibility limits: ages 35 to 85, no prior breast cancer, DCIS or LCIS, no BRCA1 or BRCA2 mutation, no chest radiation for Hodgkin lymphoma. Outside those limits the number is not meaningful.
- Already diagnosed with breast cancer? This page is the wrong tool entirely. A different test, the Breast Cancer Index, applies after diagnosis, and it is explained in the section below.
- Quick Answer
- How the Gail model actually calculates your risk
- Worked examples, checked against the tool
- Reading your numbers and the three bands
- Who should not use this calculator
- Not this tool: the Breast Cancer Index is a different test
- What most online Gail calculators get wrong
- Frequently Asked Questions
- Add This Calculator to Your Website
- References
How the Gail model actually calculates your risk
The model works in two layers. First it builds your relative risk from your answers, using coefficients fitted on large screened populations. Then it converts that into an absolute probability by running it, year by year, against the breast cancer incidence rates and the competing risk of dying from other causes for women of your age and race. That second layer is what most lookalike calculators skip, and why the same relative risk gives different absolute risks at different ages. That lifespan context is a calculation of its own: the average life expectancy calculator sketches a likely lifespan from a country baseline and lifestyle factors, which is the backdrop the lifetime risk figure runs against.
Your relative risk is the exponential of a weighted sum: coefficients for biopsies, age at first period, age at first live birth and affected relatives, an interaction between first-birth age and relatives, a biopsy-by-age interaction that weakens the biopsy effect from age 50, and a separate multiplier for atypical hyperplasia. First-birth age and family history are not independent multipliers. They sit in one joint table, published in Gail’s 1989 paper, where the combination behaves differently from either factor alone:
| Age at first live birth | 0 relatives | 1 relative | 2 or more relatives |
|---|---|---|---|
| Under 20 | 1.00 | 2.61 | 6.80 |
| 20 to 24 | 1.24 | 2.68 | 5.78 |
| 25 to 29, or no births | 1.55 | 2.76 | 4.91 |
| 30 or older | 1.93 | 2.83 | 4.17 |
Read the last column: with two or more affected relatives, a later first birth lowers the joint factor. That pattern is in the original data, and any calculator that multiplies a family-history weight by a first-birth weight gets it backwards. The other White-model factors: periods at 12 to 13 multiply risk by 1.10, under 12 by 1.21; one biopsy by 1.70 before age 50 and 1.27 from 50, two or more by 2.88 and 1.62; atypical hyperplasia by 1.82, a biopsy with no atypia by 0.93. Black, Hispanic, and Asian-American women get their own published coefficient sets (CARE study, San Francisco Bay Area Breast Cancer Study, and Asian-American study data), with some factors grouped or dropped where those studies found no effect. There is no BMI term: the Gail model does not use body weight, an honest difference from Tyrer-Cuzick.
The coefficients and rate tables here are transcribed from the published BCRA model code, the NCI model’s reference implementation on CRAN, and the joint table above reproduces Table 1 of Gail’s 1989 paper to its published rounding. The worked examples include the model’s own published test case, so you can check this implementation against the source yourself.
Worked examples, checked against the tool
Pre-filled profile. Age 50, White, periods at 12 to 13, first birth at 20 to 24, no affected relatives, no biopsies. Relative risk 1.37x (1.10 for menarche times 1.24 for first birth). Five-year risk 0.86% against an average of 1.29%; lifetime risk 8.0% against an average of 11.2%. Band: Average Risk. A relative risk above 1 with an absolute risk below average is not a contradiction: the ratio compares you with the lowest-risk answer profile, while the average woman sits above that profile by construction.
Stacked risk factors. Age 55, White, periods before 12, first birth at 30 or older, one affected relative, one biopsy with no atypia. Relative risk 4.05x at her age (5.40x under 50; the biopsy term weakens from 50). Five-year risk 3.09% against an average of 1.48%; lifetime risk 20.1% against an average of 10.2%. Band: High Risk, and the lifetime figure crosses the 20% line where MRI enters the screening conversation. The right response is a prompt appointment and a formal assessment, not alarm: the model estimates group probabilities, and her clinician can refine the picture with her full history.
Reading your numbers and the three bands
Your result has four numbers and a band. The 5-year absolute risk is the estimated percentage chance of a first invasive breast cancer in the next five years: 0.86% means roughly 9 in 1,000 women with that profile. The lifetime risk repeats the calculation to age 90. The two averages are the same calculation at population-average risk for your age and race, and they matter because a 2% risk means something different at 40, where the average is well under 1%, than at 70. The relative risk compares you only with the lowest-risk answer profile, which is why the averages are the numbers to judge yourself against.
| 5-year risk | Band | What typically follows |
|---|---|---|
| Below 1.67% | Average | Standard screening for your age, as your doctor advises |
| 1.67% to 2.99% | Elevated | A conversation about risk-reducing medication and your screening schedule |
| 3.00% and above | High | Formal assessment, enhanced screening planning, possible genetic counseling |
The 1.67% line comes from the tamoxifen prevention trials, where it defined the increased-risk group, and guidelines still use it for discussing preventive medication. A lifetime risk of 20% or higher is where annual breast MRI is typically added to mammography. For screening context, the US Preventive Services Task Force recommends biennial mammography for women aged 40 to 74 at average risk. Where your schedule sits relative to that baseline is a decision for you and your doctor, not this page alone.
Who should not use this calculator
The Gail model estimates the risk of a first invasive breast cancer, and it does not apply outside that group. Do not use it, here or on the official site, if you have already had invasive breast cancer, DCIS or LCIS; if you carry, or may carry, a BRCA1 or BRCA2 mutation or another hereditary syndrome; or if you had chest radiation for Hodgkin lymphoma. The NCI states these exclusions for BCRAT itself, because its rates and coefficients were not built from those populations. Genetic risk runs through genetic counseling and pedigree-based models instead.
Age is a hard limit too: the model was validated for women 35 and older, and this calculator accepts 35 to 85 only. Under 35 with a strong family history, that history itself is the reason to see a doctor. And no risk score outranks a symptom: a lump, skin change, or nipple change is a reason to book an appointment now, not after a calculator agrees.
Not this tool: the Breast Cancer Index is a different test
Readers sometimes arrive here looking for the Breast Cancer Index, a very different thing. This page estimates the chance of a first breast cancer in a woman who has not been diagnosed. The Breast Cancer Index is a laboratory test on tumor tissue after an early diagnosis: it measures gene activity to estimate the risk of distant recurrence, and to predict whether extending hormone therapy beyond five years is likely to help. No questionnaire can produce its score. If you have a diagnosis and a BCI report in hand, the breast cancer index calculator interprets that report’s score, nodal status and H/I result against the published risk bands. If you have not been diagnosed, that page is not for you, and this one is only a first step toward a conversation with your doctor.
What most online Gail calculators get wrong
Most online Gail calculators are a shortcut wearing the model’s name: multiply a relative risk by a single baseline percentage and call the product a 5-year risk. That skips the machinery that makes BCRAT BCRAT: no year-by-year integration, no competing mortality (what keeps lifetime risk honest at older ages), no attributable-risk adjustment, and usually no lifetime figure or average comparison. Some quote worked examples whose numbers cannot be reproduced from the inputs they state.
This page takes the opposite approach, and its limits are worth naming. Every coefficient and rate table comes from the published model code, the joint table is printed above rather than hidden, and the worked examples, including the NCI tool’s own test case, are produced by the code on this page. What it cannot do: see your mammographic density, hormone therapy use, or your relatives’ ages at diagnosis, because the Gail model uses none of those. As a Cleveland Clinic Journal of Medicine review puts it, BCRAT and the IBIS Tyrer-Cuzick model are the tools used in practice, with Tyrer-Cuzick the more comprehensive. If your risk is driven by a detailed family pattern, ask your doctor whether Tyrer-Cuzick fits you better. And this implementation is not the NCI website: for a clinical figure, run the official tool with your care team.
Frequently Asked Questions
It runs the same published model. The coefficients, race-specific baseline rates, and method here are the Gail model as published in the NCI model’s reference code (the BCRA implementation of the BCRAT SAS macro), and the worked examples show this calculator reproducing the NCI tool’s own published test case to its rounding. It is not the NCI website, the NCI has not reviewed this page, and for a clinical assessment use the official tool at cancer.gov with your doctor.
Validated models like BCRAT estimate risk reasonably well for groups of women but much less precisely for any one individual. A 2% five-year risk means that among 100 women with your profile, about 2 would be expected to develop invasive breast cancer in five years. Useful for guiding screening and prevention conversations; useless as a personal prediction or diagnosis. Simpler group measures use the same population thinking: the attack rate calculator expresses what share of an exposed group became ill during an outbreak, as a percentage of everyone at risk.
An estimated 1.67% chance, about 1 in 60, of developing invasive breast cancer in the next 5 years. It is the cutoff inherited from the tamoxifen prevention trials: at or above it, a woman is classed as at increased risk, triggering a talk about risk-reducing medication and enhanced screening. The bands here: below 1.67% average, 1.67% to 2.99% elevated, 3.00% and above high.
Commonly, a 5-year risk of 1.67% or higher, or a lifetime risk of 20% or higher. The 20% lifetime line is where annual breast MRI is typically added to mammography. If either figure crosses those lines, the next step is a planning conversation with your care team, not worry on your own.
Gail (BCRAT) uses a compact set of core factors for women 35 and older, and sits behind most medication-eligibility thresholds. Tyrer-Cuzick (IBIS) adds BMI, hormone therapy use, mammographic density, and a far more detailed family history, which suits above-average and familial risk better. Clinicians often use both.
No. BCRAT cannot accurately estimate risk for women who carry a BRCA1 or BRCA2 mutation or another hereditary syndrome, and the NCI says so explicitly. Hereditary risk assessment runs through genetic counseling, where pedigree-based tools estimate mutation probability from detailed family history. If breast or ovarian cancer runs strongly in your family, that route matters more than any questionnaire score.
No. BCRAT estimates the risk of a first invasive breast cancer and is not designed for women with a personal history of breast cancer, DCIS or LCIS. Follow-up planning belongs with your oncology team, and tests such as the Breast Cancer Index address a different, post-diagnosis question.
Never. Calculators estimate risk; mammograms and MRI detect cancer. A low estimate is not a reason to skip screening, and a high estimate is not a finding of disease. Screening decisions belong with your doctor.
Book a doctor’s appointment and take the result with you. Ask about the right screening schedule for your age and history, whether risk-reducing medication is worth discussing, whether a Tyrer-Cuzick assessment would add anything, and whether genetic counseling fits your family history. A high number is a reason to get the formal assessment done, calmly and soon, not a reason to panic.
Add This Calculator to Your Website
References
- National Cancer Institute, “Breast Cancer Risk Assessment Tool: Online Calculator” – the official BCRAT (Gail model) tool, its inputs, its validated populations, and its stated limitations for BRCA carriers and women with prior breast cancer. NCI: Breast Cancer Risk Assessment Tool. Accessed October 2026.
- Cleveland Clinic Journal of Medicine, “Breast cancer risk evaluation for the primary care physician” – BCRAT and IBIS Tyrer-Cuzick compared, including where each is and is not useful. CCJM: Breast Cancer Risk Evaluation. Accessed October 2026.
- US Preventive Services Task Force, “Screening for Breast Cancer: Recommendation Statement” (JAMA, 2024) – biennial screening mammography recommended for women aged 40 to 74. USPSTF: Breast Cancer Screening Recommendation. Accessed October 2026.
- American Cancer Society, “What We Do To Fight Breast Cancer” – about 1 in 8 US women will be diagnosed with invasive breast cancer in her lifetime, with ACS screening guidance by age. American Cancer Society: Breast Cancer Action. Accessed October 2026.
- Breastcancer.org, “Breast Cancer Risk Calculators” – how BCRAT and the BCSC calculators differ, what each includes, and the limits clinicians flag for each model. Breastcancer.org: Risk Calculators Compared. Accessed October 2026.
- University of Rochester Medical Center, “What Is a Breast Cancer Risk Assessment, and Should You Get One?” – BCRAT versus Tyrer-Cuzick inputs and the 15% and 20% lifetime bands used in screening planning. URMC: Breast Cancer Risk Assessment. Accessed October 2026.
- MD Anderson Cancer Center (patient education), “Gail Model: Breast Cancer Risk Assessment Tool” – a plain-language walkthrough of the Gail questions, its weaknesses, and when to recalculate. MD Anderson: Gail Model Patient Guide. Accessed October 2026.
- “Breast Cancer Risk Assessment Models: a brief review” – Gail and Tyrer-Cuzick model histories, input sets, validation populations and reported accuracy ranges. Review: Breast Cancer Risk Assessment Models. Accessed October 2026.
- BCRA R package (version 2.1.2), “A Package for Breast Cancer Risk Assessment” – the published R implementation of the NCI BCRAT SAS macro: the relative-risk coefficients, category recodings, baseline incidence and competing-mortality tables, and attributable-risk constants used by this calculator. BCRA: Breast Cancer Risk Assessment package. Accessed October 2026.
- MSD Manual, “Gail Model for 5 Year Risk of Breast Cancer (1999 paper)” – the model’s published relative-risk factor tables (menarche, biopsy, atypical hyperplasia), used here as an independent check on the coefficients. MSD Manual: Gail Model Calculator. Accessed October 2026.