Cancer screening occupies a strange place in public health. It is one of the great successes of modern medicine and one of its most frequent sources of overreach. The same test that saves lives in a well-selected population produces false alarms, unnecessary biopsies, and treatment for cancers that would never have caused harm in another. The question is rarely whether to screen at all; it is when to start and when to stop.
USPSTF recommendations are the most conservative widely used framework in the United States. Specialty societies often recommend earlier or more frequent screening. Neither is uniformly correct. Personal risk, family history, and preferences matter.
Cancers with strong screening evidence
Colorectal cancer. Screening now begins at age 45 for average-risk adults and continues through 75, with continuation to 85 as an individual decision. Options include colonoscopy every 10 years, flexible sigmoidoscopy every 5 years, stool-based tests annually (FIT) or every 3 years (FIT-DNA), or CT colonography every 5 years. Colonoscopy is not required; any consistent screening beats no screening.
Cervical cancer. Begin at 21 with Pap testing every 3 years. From 30 to 65, options expand to include HPV testing alone every 5 years or co-testing every 5 years. Screening can stop at 65 with adequate prior testing and no history of high-grade lesions.
Breast cancer. USPSTF now recommends biennial mammography from 40 to 74 for average-risk women. Women with dense breasts, family history of premature or ovarian cancer, or known genetic mutations may need earlier, more frequent, or supplemental screening with ultrasound or MRI. Screening beyond 74 is a shared decision.
Lung cancer. Annual low-dose CT for adults 50-80 with a 20 pack-year smoking history who currently smoke or quit within the past 15 years. This is the most under-utilized high-yield screening in adult medicine. If you are eligible and have not been offered it, ask.
Cancers where screening is nuanced
Prostate cancer. USPSTF recommends shared decision-making about PSA testing for men 55-69, with individual choice earlier for Black men or those with a strong family history. Not recommended after 70 in most cases. The trade-off is real: PSA testing reduces prostate cancer mortality modestly and produces meaningful overdiagnosis.
Skin cancer. USPSTF finds insufficient evidence for routine whole-body screening in average-risk asymptomatic adults. Dermatologic surveillance is reasonable for people with fair skin, many moles, family history of melanoma, or immunosuppression. Self-examination for new or changing lesions is universally reasonable.
Ovarian, pancreatic, and testicular cancer. Not recommended for routine screening in average-risk adults. Focused surveillance may be appropriate for people with genetic risk (BRCA, Lynch, Peutz-Jeghers, and others), managed by specialists.
When to stop
The stopping conversation is under-taught and often deferred. USPSTF explicitly notes that most cancer screenings should end when life expectancy drops below roughly 10 years, because the benefits of finding disease early diminish while the harms of workup and treatment do not.
For adults in their late 70s and 80s, screening decisions should account for functional status, comorbidity, and personal preferences. A vigorous 80-year-old may reasonably continue mammography; a frail 72-year-old may reasonably stop. This is a conversation, not a formula.
Screening is a service, not a rite. It should be offered when it helps and withheld when it does not.
Where family history changes things
A first-degree relative diagnosed with colorectal, breast, prostate, or ovarian cancer at a young age typically shifts screening earlier, sometimes by a decade or more, and may add MRI or genetic testing to the plan. Known pathogenic mutations move people out of general population guidelines and into specialty-managed surveillance.
The bottom line
Start cancer screening on time, use the modality you are most likely to actually complete, and revisit the plan every few years. When life expectancy or quality of life changes, so should the plan. The right answer is not always more.