I'm wondering what sorts actual values we can expect to find for peculiar velocities of individual objects when observing the universe, both typically and on the high end. I'm interested in any reasonable definition of peculiar velocity — though I'm generally thinking about it relative to the Hubble flow.

I've seen numbers around a few hundred km/s for things like the motion of the solar system relative to the local group or the CMB, or the local group relative to the great attractor. I've also found galactic rotation curves upwards of a few hundred km/s. And the measurement of the sun's motion relative to the local standard of rest has individual stars around the same order of magnitude. So I'm gonna say bog-standard peculiar velocities may get up to about 1,000 km/s. Is that reasonable? Is there a common class of objects that will typically have higher velocities?

Now, at the high end, I would expect that the very highest velocities involve close approaches to black holes. Are there other mechanisms that might push objects to higher speeds reasonably often? I see that one star (S62) orbiting Sgr A* reaches about 30,000 km/s on closest approach. Is this the highest speed ever measured in the universe?

  • 1
    $\begingroup$ potentially related: What does “unremarkable transverse peculiar velocity” mean exactly, and how is it calculated here? $\endgroup$
    – uhoh
    Aug 19, 2021 at 0:52
  • $\begingroup$ Meh. I guess it's related, but the whole problem is that I understand why peculiar velocities happen, but I don't know the actual values that have been observed. I've had a really hard time finding anything in the literature that gives actual numbers. $\endgroup$
    – Mike
    Aug 19, 2021 at 18:45
  • $\begingroup$ "...but I don't know the actual values that have been observed. " You've accepted pela's answer which means you're satisfied that your question has been completely answer and don't need anything else. If you need a larger number of values and the links in the answer are not what you need, then you should express that to them and at least temporarily un-accept his answer. $\endgroup$
    – uhoh
    Aug 19, 2021 at 21:56

1 Answer 1


From the virial theorem, peculiar velocities resulting from gravitational processes generally have magnitudes of the order of $$ V \sim \left(\frac{GM}{R}\right)^{1/2}, $$ where $G$ is the gravitational constant, and $M$ and $R$ are the mass and radius of the system.


Galaxies have peculiar velocities of the order of several 100 km per second; small velocities for small groups (~100 km/s; e.g Carlberg et al. 2000) and large velocities for rich clusters (~1000 km/s; e.g Girardi et al. 1993).

The maximum velocities are reached when a galaxy plunges through the center of the cluster, and many clusters thus have a few galaxies surpassing 2000 km/s (e.g. Einasto et al. 2010).


Many galaxies eject particles through galactic winds, powered by active galactic nuclei (AGN) and/or stellar feedback (radiation pressure and cosmic rays from massive stars, as well as kinematic and thermal feedback from exploding supernovae). Although these processes are not gravitational per se, the outflow velocities typically do scale with the size of the galaxy due to various scaling relations (more massive galaxies host more massive central, supermassive black hole, and have more vigorous star formation).

The particles don't always escape; typically the outflow velocities $v_\mathrm{out}$ are a bit smaller than the galaxies' circular velocities $v_\mathrm{circ}$ which, in turn, are a factor $\sqrt{2}$ smaller than the escape velocity $v_\mathrm{esc}$. For instance, Rupke et al. (2005) find $v_\mathrm{out} \simeq v_\mathrm{circ}^{0.85}$.

Hence, typical outflow velocities of gaseous material are several 100 km/s (e.g. Sugahara et al. 2017), but sometimes surpass 1000 km/s and escape (which is the reason that the intergalactic medium has been enriched with metals).

However, the most highly accelerated parts of the jets resulting from AGN often reach relativistic velocities, i.e. velocities approaching the speed of light $c$. Similarly, cosmic rays and neutrinos have velocities which are very close to $c$. These are, of course, the maximum peculiar velocities you'll find, but I suppose you're thinking of macroscopic bodies, not particles.


Stellar velocities — that is, the velocities with which stars move around inside galaxies — also increase with the mass of the galaxy. In a galaxy like the Milky Way, a fairly typical galaxy of $M\simeq10^{12}\,M_\odot$, stars revolve with velocities of 200–250 km/s, with a dispersion of a few tens of km/s (that is, in the reference frame of a given star, nearby stars typically have velocities of order ~10 km/s.

However, sometimes stars in dense environments (globular clusters) come close to each other and acquire larger velocities. And if binary stars come close the the galaxy's central black hole, it can reach very high velocities, more than 1000 km/s.

The record for such hypervelocity stars is 1755 km/s (Koposov et al. 2019).

Asymmetric supernova explosion in stellar binaries may also give rise to runaway stars with velocities of several 100 km/s (Tauris & Takens 1998).

Solar system bodies

For the sake of completeness, I'll end by discussing planets, though they of course have smaller peculiar velocities (in the reference frame of their star). You can use the same formula as given in the top, and since most of the mass is concentrated in the center (the star), $M$ and $R$ are just the mass of the star, and the distance between the star and the planets, respectively. For instance, Mercury with a distance of 0.4 AU from the Sun has a velocity of 47 km/s.

Sometimes planets are ejected and become rogue planets, which may reach velocities of more than 100 km/s (Bardalez Gagliuffi et al. 2020).

  • 1
    $\begingroup$ Thanks. The Einasto and Koposov references look like just what I wanted. $\endgroup$
    – Mike
    Aug 19, 2021 at 18:45
  • $\begingroup$ @Mike I realize now that you're probably mostly interested in stellar velocities near black holes, so don't know if all the stuff I wrote about galaxies is relevant to you. But if so (you do mention that the Einasto ref may be useful), there are several surveys with velocity histograms in clusters, showing the dispersion. And yes, I think S62 (Peißker+ 20) holds the record for the highest-velocity star, at $0.1c$ (pretty crazy). $\endgroup$
    – pela
    Aug 19, 2021 at 21:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .