In July 2022, 48% of U.S. adults told Pew that online dating was very or somewhat safe, and 49% said it was not too safe or not at all safe.
On this page
What adults meant when they rated safety
Pew Research Center asked U.S. adults in July 2022 whether dating sites and apps were a safe way to meet people. Forty-eight percent said very or somewhat safe, while 49% said not too or not at all safe. These responses describe public judgment about a way of meeting. Pew did not follow each respondent through a date, count every harmful event, or test a safety feature. The nearly even split is a real finding about opinion, and it deserves to keep the word 'opinion' beside it. A person can believe a service is generally safe and still take precautions.
The two sides do not quite total 100%, because some responses were unreported or affected by rounding. It would be wrong to turn the remaining few points into a third group with a specific view. Pew also found that 53% called online dating safe in its 2019 survey, compared with 48% in 2022. That is a change in survey responses across two points in time. It does not identify why views changed, and it does not measure a matching change in harm. Read the dates with the figures before applying them to the present day.
Three different safety questions
The table may be wider than the screen. Focus it and use the arrow keys to reach every column.
| Question | Published result | People counted |
|---|---|---|
| Is online dating safe? | 48% safe; 49% unsafe | All U.S. adults |
| Is it safe, by use history? | 67% of users; 41% of nonusers said safe | Each past-use group |
| Require background checks? | 60% support; 15% oppose; 24% unsure | All U.S. adults |
Key distinction: A view about safety, a view by past use, and support for a rule answer different questions.
How to read it: Pew did not test whether background checks prevent harm. Percentages may not total 100 because of rounding or unshown responses.
Download the table (CSV) ↓Past users and nonusers answered differently
Among adults who had tried online dating in Pew's 2022 survey, 67% said it was safe to meet people this way. Among adults who had never tried it, 41% said so. The question was the same, but the people answering had different use histories. The gap is descriptive. It cannot tell whether trying a service made users feel safer, whether people who already felt safer were more likely to try it, or whether other differences between the groups shaped their answers. A headline that calls the user figure the national view quietly leaves out the larger group of nonusers.
Safety views also differ within the user group. Among women who had ever used a site or app, 41% called online dating unsafe, compared with 23% of men who had used one. That comparison belongs to past users, rather than all women and all men in the country. It does not say what happened to each respondent or identify a particular service. The gender difference is worth taking seriously as a reported difference in judgment, while the separate unwanted-behavior questions are needed to describe specific experiences.
A safety opinion and an unwanted incident are separate counts
Pew's safety question asked whether online dating seems a safe way to meet people. In a different part of the 2022 survey, 48% of past users reported at least one of four listed unwanted behaviors on a dating site or app. That second denominator excludes people who had never used online dating, and the question concerns events rather than general perception. The two 48% figures are numerically equal by coincidence of rounding and should never be substituted for one another. One does not reveal how many of the people who reported harassment also rated the platforms unsafe.
The listed behaviors range from unwanted sexual messages or images to threats of physical harm. They vary in severity and can occur more than once, yet the combined survey measure counts a person if at least one happened at any time while using online dating. It is therefore neither an incident rate per date nor a count of crimes. For a specific behavior, read the harassment results. For public confidence, read the safety-opinion result. The safety table on this page helps keep the questions apart, which is more useful than compressing them into one claim that apps are simply safe or unsafe.
What support for background checks says
Pew also asked whether dating companies should require a background check before someone creates a profile. In 2022, 60% of all U.S. adults favored that requirement, 15% opposed it, and 24% were unsure. This is support for a proposed rule, not an evaluation of a check that was put into practice. The survey cannot establish whether checks would prevent harm, what records they would find, or how people would use the results. A clean search result cannot establish that a stranger is safe to meet.
The full-population figure also hides a divide by experience. Pew reported support from 47% of past users and 65% of adults who had never used online dating. Those groups may imagine different costs and benefits, but the survey does not prove why they answered differently. The opinion result can inform a policy discussion; it cannot promise a protective effect for an individual meeting. If you are looking at an app's current safety claim, ask what its check actually covers, when it runs, and whether the service explains its limits. Do not read the 60% as a guarantee of any feature.
Support also varied by age in Pew's 2022 survey: 53% of adults under 50 favored required checks, compared with 62% of ages 50 to 64 and 73% of adults 65 and older. These are within-age opinions, not evidence that a check helps one age group more. A proposal can attract support for many reasons, including concern about meeting strangers, yet the survey did not ask each supporter to choose a reason. Quoting the all-adult 60% alone hides that spread. If a policy claim uses the age pattern, keep the age groups and the fact that the question asked about a requirement before profile creation.
A later survey asked about meeting in person
SSRS asked U.S. adults a related but different question in January 2026: whether meeting someone in person after connecting on a dating site or app was generally safe. Fifty-seven percent said safe and 43% said not too or not at all safe. SSRS also found that 55% of women and 30% of men considered that meeting unsafe. The wording focuses on an in-person meeting, whereas Pew's 2022 item asks whether dating sites and apps are a safe way to meet people. The dates, samples, and wording differ. Do not read 48% and 57% as a measured rise in the same opinion.
The 2026 result still adds useful context: a person may distinguish between browsing a platform and meeting a person in the physical world. The SSRS figure is a view about that transition, not a count of encounters or a test of precautions. It should be named with its question and year. If someone cites only 'most people say online dating is safe,' ask whether they mean Pew's platform question, SSRS's meeting question, or another survey altogether. Those details determine what the percentage can fairly support.
Use safety guidance without shifting responsibility
RAINN's safer-dating guidance offers concrete options, including meeting in a public place, arranging your own transport, and blocking or reporting troubling behavior through a service. A person can use those options when they fit the situation. No step can control another person's conduct, and responsibility for harassment or violence stays with the person who commits it. The survey figures do not tell you how a particular stranger will act. They are best read as evidence that concerns differ across groups and that harmful experiences are common enough to deserve serious attention.
For a claim about how safe people feel, keep Pew's 2022 48%/49% split and the all-adult denominator. For a claim about what users have faced, go to the behavior questions. If you are planning a meeting, use current safety information from the service and trusted guidance rather than a national opinion share as a personal clearance. The FTC warning-sign guide linked from this page addresses a different risk: fraud and requests for money. Each resource answers its own question, and together they give you a clearer basis for action.
