Count me skeptical about reports on the number of Catholics in the world and in the United States. When Pope Francis died, Reuters and Al Jazeera both reported there were 1.4 billion Catholics world-wide. CBS News reported 1.3 billion. The World Christian Database reported 1.272 billion. The Italian-language Vatican Annuario Pontificio (Pontifical Yearbook) 2022 gave the number of Catholics in the world at the end of 2020 as 1,359,612,000. For the end of 2021, that publication used 1.376 billion. So, these six figures, extrapolated to 2025 and/or rounded, are within 100 million of each other. Too large a spread, eh?

Where do these numbers come from? I can answer with respect to the numbers from the Annuario Pontificio and its related volume Annuarium Statisticum Ecclesiae, which are compiled by the Central Office for Statistics of the Church, part of the Vatican’s Secretariat of State. According to the Annuario Pontificio 2012, the statistical data given in these yearbooks regarding archdioceses and dioceses are furnished by the diocesan curias concerned and reflect the diocesan situation on December 31 of the year prior to the date on the yearbooks, unless there is another indication. How reliable is this method of relying on the dioceses?

Counting Registered Parishioners

At least for the United States, the numbers from most, or many, dioceses come from the numbers of registered parishioners. Each parish’s number is rolled up into a diocesan number. What are the numbers for the United States?

You may be surprised, as I was, to learn that the annual Official Catholic Directory, published by P.J. Kenedy and Sons with information and statistics for every diocese in the United States, does not add the numbers and provide a number for the entire United States. Furthermore, while the webpage for the United States Conference of Catholic Bishops (USCCB) has some statistics for various populations like Blacks and Hispanics, even it does not set forth the national statistics. Moreover, Our Sunday Visitor’s last Catholic Almanac was 2016.

I would assume the numbers for the United States would appear in the Vatican’s Annuario Pontificio, but I’m not sure. I have not purchased a copy and it is no longer available at the University of Virginia library (which stopped purchasing the book in 2019). Furthermore, news reports issued when a new annual Annuario Pontificio is released describe the global and regional figures but not those of any individual countries, including the U.S. (e.g., “New Church Statistics Reveal Growing Catholic Population, Fewer Pastoral Workers,” Vatican News Agency, March 2025; republished by Catholic News Agency, March 21, 2025, and by National Catholic Register, March 21, 2025).

I have found one report that published a number for the United States which used the Official Catholic Directory for its source. In 2022 a group of specialists in church statistics worked together to produce Catholics in the U.S. Religion Census (Clifford A. Grammich, Nov. 2022). In this report, Grammich states, “Diocesan totals tabulated from The Official Catholic Directory for 2020 indicate there were 67,635,546 Catholics in the United States that year.” Note that he doesn’t say that the Vatican Annuario Pontificio provides a number. One would think that if it did, he would use it. He writes that the number for 2020 was “tabulated” from the Directory. Someone associated with him took the Directory‘s figures for each (arch)diocese and added them up.

Whatever the number for whatever the year, there are huge methodological issues with the Official Catholic Directory’s numbers for the (arch)dioceses. These are described, in part, by Grammich as follows:

Dioceses are inconsistent in their reporting to the 2020 directory. Some dioceses, for example, appear to guess a population percentage and apply that to their assumption of the total population—and are often wrong both in the population percentage they guess and the total population they assume. Others use [parish] registration statistics, but these understate the total Catholic population, especially in rapidly growing areas. Still other dioceses report identical totals for several consecutive years, regardless of other population change in their area.

Let me expand on Grammich’s concerns about the numbers (arch)dioceses report. First, you are no doubt familiar with the various issues concerning the purging of voter rolls of voters who have died or moved out of the jurisdiction. The same issue occurs with respect to registered parishioners. There is no process throughout the United States for purging the names of those who have died or who are otherwise no longer members of the parish. Indeed, people can be registered in more than two parishes. I personally have sometimes, not always, told a priest or staff member I was moving, but I never asked them to follow through and disenroll me. My current diocese, the Diocese of Richmond [Virginia], uses software that removes names from their former parish registrations if they register in a new parish — but only if they do so in a parish within the diocese. If they register in a new parish outside the Diocese of Richmond, they remain on the rolls of the parish inside the Diocese. Thus, whatever number is provided as the total number of registered parishioners in the United States, it could be millions, even tens of millions, over the actual number.

A second issue with relying on the number of registered parishioners in each (arch)diocese to generate a national number is the problem that the number of registered parishioners can understate or overstate the actual number of Catholics. It can understate the actual number because there is no requirement for a Catholic to register with any parish. (The failure to register can make it difficult to have your children attend the parochial school, or marry, or be visited in the hospital, or have a funeral Mass, but there is no requirement.) So, there may be many Catholics, practicing or not, who are not registered in any parish. It can overstate the actual number because a registered parishioner may not regularly attend Sunday Mass, or indeed may not attend any Mass (except perhaps Christmas and Easter) but still wants the benefits of registration.

(For similar and additional critique, see Michael Cieslak, “The Best Way to Estimate the Catholic Population,” The Catholic Conversation: From a Sociologist’s Perspective, May 28, 2012, and  “Counting Catholics: A Comparison of 3 Methods,” June 4, 2012.)

We can say this about the American Official Catholic Directory and the Vatican’s Annuario Pontificio: While their numbers may not be precise, the method of computation arguably uses apples and apples from year to year and might show statistically valid trendlines. Of course, to conduct various statistical analyses requires that datasets be made available for researchers. Apparently they aren’t. According to a just published book, Global Catholicism: Between Disruption and Encounter (by Massimo Faggioli and Bryan Froehle, 2025), researchers need to create their own datasets:

Sean [C. Thomas] has…pioneered [since 2021] a number of new approaches to global Catholic statistics. He devised ways to scan and convert the pages of the [Vatican’s] Annuarium Statisticum Ecclesiae into electronic datasets after extensive tests to assure data quality. He did similar work with the Annuario Pontificio and other Catholic data sources.

That is, Thomas had no choice except to create his own electronic datasets.

Sampling: An Alternative to Counting Registered Parishioners

One alternative to counting, and rolling up, the number of registered parishioners to derive the number of Catholics in the United States or world-wide would be to ask a question about religion through a census. In the United States, however, the law forbids the Census Bureau from asking about religious affiliation, leaving the work to be done by social scientists on a private basis.

Another alternative to using the number of registered parishioners is sampling or surveying, whether by phone or door-to-door or some other method. (No doubt you are familiar with this when it is done for political polling.) Once a percentage of Catholics in a national sample has been quantified, the number of Catholics nationally can be obtained by multiplying by the number of people in the general population.

Defining “Catholic” for Polling

Before proceeding, let me observe that too often political polls conducted in the United States do not define “Catholic.” Generally the polls are referring to people who self-identify as Catholic. Sometimes they will distinguish these from practicing Catholic (that is, those who attend Sunday Mass regularly). Thus, an article by Tyler Arnold in the November 6, 2024, edition of National Catholic Register (“Catholic Voters Favor Trump Over Harris Nationally and in Swing States, Exit Polls Show”) describes three exit polls in the November 2024 presidential election seeking to assess how Catholics voted. The article cites polls published by the Washington Post, the Associated Press, and NBC News. There was no mention in any of these polls of how Catholics were defined. (Let me add that neither the link given in the article for, nor my research on, the Associated Press VoteCast poll showed any mention of Catholics. Also, the NBC News poll was limited to ten states.)

I’ll also observe that while a “practicing” Catholic might be defined as one who attends weekly Sunday Mass, another criterion could be belief in the Real Presence. A CARA report of 2023, cited below, found that 95% of those who attend Sunday Mass do so believe, but only 38% of all those who identify as Catholic know and accept this. That is, most of those who identify as Catholic but do not attend weekly Sunday Mass do not believe (generally they do not know the doctrine, so do not reject it).

Results of Various Polls

A Pew Research Center poll of over 9,500 adults — limited to the United States — was taken in February 2025. Twenty percent identified as Catholic. (Nine percent more said they are former Catholics.) Among those who identified as Catholic, 28% attend weekly Sunday Mass. So, extrapolating from the general population, both adults and children, estimated to be 340 million in 2025, there are 68 million residents of the United States who identify as Catholic, 19 million of whom attend Sunday Mass.

The Center for Applied Research in the Apostolate (CARA) of Georgetown University also conducts polls, limited to the United States. CARA recently stated: “Since 2000, CARA has been surveying Catholics nationally. CARA conducted 43 of these surveys from 2000 to 2024. CARA has utilized interviewing by telephone and self-administered surveys onscreen. CARA can conduct polls with probability-based sampling or using opt-in online panels. The variability in sampling and survey methods allows for the widest range of clients to commission surveys of the Catholic population.”

One such CARA survey was published in September 2023 by Mark M. Gray, entitled “Eucharistic Beliefs: A National Survey of Adult Catholics.” The number of self-identified Catholics polled was 1,031. The report noted (in Note 2): “CARA Catholic Polls consistently estimate that between 22 percent and 23 percent of the adult population in the U.S. self identifies as Catholic.” (By comparison, the Pew report I cited above said 20% identify as Catholic.) The CARA report stated 24% of those who self-identify as Catholic attend weekly Sunday Mass. (By comparison, the Pew report said 28% did.)

In Grammich’s November 2022 report, Catholics in the U.S. Religion Census, his team asked each diocese for numbers. The team also used the Official Catholic Directory and survey statistics, “particularly a special Pew [Research Center] file made available.” The resulting number the team computed was 61,858,137 for 2020.

Grammich therefore listed the following numbers:

76,600,000 (in 2020) (National Opinion Research Center General Social Survey, NORC-GSS, which found 23.13% in 2018, applied to the national population in 2020)

74,200,000 (in 2020) (Nationscape, which found 22.3% in 2019-2020)

61,858,137 (in 2020) (Catholics in the U.S. Religion Census)

67,635,546 (in 2020) (Official Catholic Directory, per Grammich)

To these four numbers, I add the fifth, from Pew:

66,200,000 (in 2020) (Pew found 20% or 68 million in 2025; I applied this percentage to the national population in 2020 to compare apples to apples with the other numbers)

So, for 2020, these various attempts at quantifying the number of Catholics in the United States yielded results ranging from 61.9 million to 76.6 million. The highest figure is 23.75% more than the lowest figure. This is a big range, both in absolute numbers and statistically speaking. Is one more statistically valid than the others?

Counting Attendees at Sunday Mass

Still another method of quantifying the number of Catholics is to count the number of people attending Mass on Sundays. Of course, this yields a number only for practicing Catholics, not all self-identified Catholics.

About six years ago, the Diocese of Arlington (Virginia) started conducting this count every year on the Sundays (and Saturday vigils) in the month of October. A January 2025 report stated that the previous October’s count was 125,164, or 28.8% of an estimated 433,000 registered Catholics in the diocese (Staff Report, “Annual Diocesan Report Indicates Mass Attendance Grows Slow But Steady [still short of pre-pandemic levels], Arlington Catholic Herald, Jan. 30, 2025). This report added: “With nearly three out of 10 registered Catholics attending Mass weekly, the diocese is slightly ahead of recent national rates of attendance. A Gallup survey last year [Jeffrey M. Jones, “Church Attendance Has Declined in Most U.S. Religious Groups,” Wellbeing, March 25, 2024] indicated 23 percent of Catholics attend Mass weekly, while Georgetown University [CARA] put the figure at 17 percent in 2023.”

(The same Gallup survey stated that Gallup “measures church attendance and religious affiliation on nearly every U.S. survey it conducts. These results are based on aggregated data from Gallup telephone surveys conducted in 2021, 2022 and 2023…The combined 2021-2023 data comprise interviews with more than 32,000 U.S. adults and at least 200 respondents in each religion…Gallup also constructed similar aggregates using 2000-2003 and 2011-2013 data to assess changes in religious service attendance over time.”)

What a huge range in Mass attendance! 28.8% in the Diocese of Arlington, 23% nationally per Gallup, 17% nationally per CARA.

Conclusion

Yes, you can count me skeptical about reports on the number of Catholics in the world and in the United States.