I am working on an introductory presentation on variable stars aimed at beginners and would like to give some basic background information. In particular I would like to give a rough estimate of the number of variable stars currently known in the Milky Way. As expected there are wildly different estimates for this around the internet, varying from 50000 to 150000, perhaps even more. Checking VSX for this information through the web interface seems to overload the database as it tries to retrieve the information on each star. I wonder if someone here with more privileged access to VSX could be so kind as to retrieve this information, namely: the number of confirmed variables in the Milky Way currently in VSX, if possible (but not necessarily so) grouped by variability type.
People with access to and control of the database can give you exact numbers. But you can find some basic information from VizieR.
For example, the catalog page for VSX on that site says it has 10,297,173 rows, which matches nicely with the sentence on the AAVSO page for the catalog that says over 10.2 million. You can also search by type on VizieR, so you could put “RR*” (without the quotes) in the Type search box and it will return all stars marked as being some kind of RR Lyrae (this can get tricky for large values, you may want to get the results as a CSV or FITS file instead of HTML).
Another good source, and possibly quicker way to get what you want, is the variability catalog from Gaia DR3. You can see from that link that they have individual catalogs for different types and it tells you how many rows (i.e. stars) are in each one. For example, there are 271,779 stars in the RR Lyrae catalog.
Hope that helps! And I’m sure someone involved with VSX directly will post something soon to answer your question.
You sent that question to the VSX email so I was working on a reply yesterday doing some queries and sent it to you a while ago.
As I said in my email, the number of variable stars in VSX is not the number of known variable stars, because lots of catalogs and lists haven’t been incorporated to VSX yet.
A guesstimate of the currently known number of variable stars (not includng suspects) should be in the order of ~15,000,000.
Rotational (most stars have some kind of spots), eclipsing binaries (they tend to come in pairs!) and pulsating variables (with so many different types) are the Top 3, with over 2 million objects each out of the 10 million in VSX.
I have now received a kind and detailed answer from Sebastián Otero. Below I post the table I received so the community can also use this data.
As you all understand, these data are only an indication of the current knowledge and should be used with caution. There are many complexities associated with this table, which Sebastián pointed out to me, for example referred to stars that possibly belong or are suspected to belong to more than one type, possible classification errors, multiple stars, etc. Also it must be noted that that these numbers where retrieved today (2026-01-28), representing today’s state of the database, and will inevitably change with time.
In any case, even with all these caveats, I find the data extremely interesting and I believe others will do as well.
Thanks again to Sebastián and the AAVSO.
Hernán
Rotational variables 3,165,181
Eclipsing binaries 2,620,781
Pulsating variables 2,089,564
RR Lyrae stars 310,811
Mira stars 88,712
Cepheids 4,535
Cataclysmic variables 17,162
Dwarf novae 11,842
Cataclysmic variables 3,450
Novae 736
Supernovae 595
Symbiotic variables 539
Eruptive variables
Young stellar objects 86,191
Be variables 2,618
Exoplanets 1,537
X-ray variables 196
Is Sol, a variable star with a roughly 11-year period superimposed with an erratic variability on the minutes to weeks timescale, in the list? It is very easy to observe, being at magnitude -27 or so.
