Can Primordial Black Holes Be All the Dark Matter?

Primordial black holes (PBHs) are constrained by evaporation, lensing, compact-binary populations, accretion, dynamics, and structure, but those constraints do not combine into a model-independent yes-or-no answer. The proposition “PBHs are all the dark matter” refers to a complete mass, spin, spatial, and formation distribution with total present fraction one—not to a single monochromatic point on an exclusion plot. Through 2026-08-20, broad and stellar-mass all-dark-matter populations face strong independent constraints, while an asteroid-scale sensitivity gap prevents a universal exclusion. No observation establishes a primordial population, so surviving parameter space is not positive evidence.

Helpful background. Primordial-Black-Hole Observational Constraints explains the physical mechanisms behind evaporation, lensing, dynamical, accretion, and gravitational-wave bounds. This assessment asks whether one physically generated population can satisfy all applicable constraints simultaneously.

The all-dark-matter existence question

The finite proposition is whether there exists a physically generated PBH distribution whose present abundance satisfies $f_{\rm PBH}=1$ while every applicable, non-double-counted likelihood and formation-consistency condition is met under defensible nuisance assumptions. A distribution must be normalized, including tails produced by critical collapse or thermal history, and evolved from formation to the present.

The answer depends on monochromatic versus extended mass functions; formation versus present abundance; Hawking emission and possible new particles; spin and accretion history; smooth versus clustered spatial distribution; lens source size, wave optics, and halo velocities; compact-binary formation and merger history; Gaussian versus non-Gaussian primordial tails; and whether published curves represent robust likelihoods, forecasts, or contested translations.

Assessment at 2026-08-20	Meaning
Question status	Open: neither $f_{\rm PBH}=1$ nor a universal exclusion is established for every physically admissible distribution
Evidence status	Insufficient evidence: a remaining allowed region is not detection evidence, while broad and stellar-mass scenarios are strongly constrained
Confidence	Moderate; high for the conditional constraint framework, moderate for the robustness and precise extent of the remaining sensitivity gap
Domain	PBHs across evaporation, lensing, gravitational-wave, dynamical, accretion, and structure probes under explicitly specified population assumptions
What could change the conclusion	A new asteroid-region constraint, a material revision of a dominant bound, or a physical extended-mass synthesis that closes or reopens $f_{\rm PBH}=1$

A starting project: find which mass-function tail matters

Choose this question after working through Combining PBH Constraints. You should be able to distinguish a density per unit mass from one per logarithmic mass and to map a formation distribution to the population tested by a constraint. Read Carr et al. 2017, §III, Eqs. (2)–(5) and (12), revised v3 in that order: normalization and shape come before exclusion. The old paper supplies a method benchmark, not today’s allowed windows.

Make the first project deliberately narrower than deciding whether PBHs are all dark matter. Select one present-day lognormal population and two individually documented linear constraints whose mass domains sample different tails. Resolve the complete normalized distribution on successively wider mass grids and record the contribution from each mass interval to each predicted signal. Compare with the monochromatic limit. If you truncate or renormalize a tail, show the change explicitly; do not interpret an integration boundary as a physical formation cutoff.

The deliverable is a notebook and two response-contribution plots explaining which tail controls each conditional limit. This needs a laptop and access to the constraint inputs; a physically generated replacement distribution needs formation expertise and may require collapse calculations. A possible next project is to replace the phenomenological mass function with one derived from a fixed formation model and test whether the same tail conclusion survives. Establish novelty through the formation and constraint literatures before proposing that extension.

Stop claiming an all-dark-matter result if the normalization refers to a different epoch, a decisive response is nonlinear, or the relevant likelihood is unavailable. The smaller project can still answer a useful question about sensitivity to tails. Even a distribution that passes every implemented test is an allowed example within those assumptions; it does not show that a primordial population exists.

What the current constraint landscape establishes

The current broad synthesis separates evaporation, lensing, gravitational-wave, dynamical, accretion, and structure probes and gives their population assumptions Carr et al. 2026, §§II, IV–V, and VIII. Its baseline retains an asteroid-mass interval that present searches have difficulty probing. That is an instrument- and physics-dependent sensitivity gap, not a theorem: Hawking emission, compact-object capture, wave optics, source sizes, clustering, and extended tails can move its edges.

Extended distributions cannot be tested by checking only their peak. Under approximately linear, separable responses, a monochromatic upper limit can be integrated against the mass function; Carr et al. derive the procedure in §III, Eqs. (8)–(12), and examples in §§III–IV 2017, §III, Eqs. (8)–(12), revised v3. The corrected Fig. 1 must be used. Merger, accretion, and structure observables can respond nonlinearly or share data, so the simple integral prescription is not universal.

Formation itself generates consequential width and evolution. Threshold profiles, critical collapse, thermal history, and non-Gaussian statistics can alter abundance exponentially and populate mass-function tails Escrivà, Kühnel, and Tada 2022, §§2–10. A hand-selected narrow log-normal curve is therefore not automatically a physically realized solution.

The newest gravitational-wave row further constrains solar-mass populations. An O4a analysis jointly models resolvable PBH mergers, an astrophysical population, and the associated stochastic contribution; the resolvable-merger inference dominates its bounds Andrés-Carcasona et al. 2026, §§2–4 and conclusion. It is a current preprint and remains conditional on PBH binary formation and merger history.

Results and open questions at 2026-08-20

Claim or evidence	What it supports	What it does not establish
Current multi-probe review and tables	A class-separated map of present constraints and a difficult asteroid region	A raw joint likelihood or theory-independent open window
Extended-mass recast formalism	Conditional treatment of tails under linear independent responses	Nonlinear merger, accretion, or correlated-data combination
Formation and critical-collapse calculations	Physically motivated widths and sensitive abundance mappings	A uniquely selected mass function or observed normalization
O4a PBH population analysis	Strong current solar-mass constraints under stated binary and astrophysical models	Exclusion of asteroid masses or proof that no merger is primordial
Evaporation and messenger constraints	Strong low-mass sensitivity under specified particle and propagation physics	Model-independent limits under arbitrary new particles or relics
No confirmed primordial identification	Absence of positive evidence for $f_{\rm PBH}=1$	Exclusion of every undetected population

Independent physical mechanisms deserve more weight than multiple curves derived from the same data or halo model. At the same time, combining them requires a population whose tails and evolution are consistent across every response. A visually closed envelope is not itself a joint confidence statement.

The strongest exclusion and survival arguments

The exclusion argument is cumulative. Broad or stellar-mass distributions that reduce one lensing signal often place abundance into evaporation, accretion, merger-rate, or dynamical regimes. The 2021 critical review details the dependence of these classes on astrophysical assumptions and extended functions Carr et al. 2021, §§3–9. Newer O4a data strengthen one solar-mass row but do not remove the population-model dependence.

The survival argument is narrower: with standard compact-object and evaporation assumptions, a sufficiently concentrated physically generated distribution may remain in the asteroid-scale sensitivity gap. Its tails must still be checked against adjacent constraints. Moreover, “allowed” means the current data lack exclusion power; it does not increase the probability that PBHs formed or supply evidence that they are dark matter. The current gravitational-wave review likewise emphasizes that compact-object identity requires population discriminants, not merely mass overlap Franciolini et al. 2024, §§2–7.

Counterexamples, null results, and failed routes

An extended distribution whose peak lies in an allowed monochromatic interval but whose low-mass tail evaporates is a counterexample to peak-only reasoning. A stellar-mass population that satisfies one microlensing curve yet overproduces mergers or accretion signatures is another. Conversely, a null microlensing search is not evidence against populations outside its cadence, wave-optics, source-size, and halo-velocity domain.

The most common failed route is to sum digitized inverse upper limits from heterogeneous plots without checking response linearity, confidence conventions, covariance, and nuisance choices. Another is to infer primordial identity from a compact-object mass. Astrophysical black holes occupy overlapping mass ranges, so a population-level formation signature is required.

What would change the assessment

A robust closure of the asteroid region would require a sensitivity calculation and likelihood whose wave optics, finite source, foregrounds, spatial distribution, and particle assumptions are public and applicable to the full mass function. A defensible positive shift would require at least two independent signatures—such as a merger-population feature and a formation-linked early-Universe signal—predicted by the same normalized distribution. A material correction to an evaporation, capture, lensing, accretion, or O4a constraint could reopen or close tails and triggers immediate review.

Research checklist

• Define the formation-time distribution and map it to present mass, spin, abundance, and clustering.
• Normalize the full function to $f_{\rm PBH}=1$ and resolve tails over every constrained decade.
• Pin the Hawking particle model, greybody factors, secondary spectra, cosmology, and propagation backgrounds.
• Recompute lensing efficiency with cadence, source size, wave optics, halo density, and velocity distribution.
• State compact-binary formation, disruption, accretion, and merger-history assumptions.
• Build a per-constraint domain mask, nuisance record, confidence convention, and overlap graph before combination.
• Perform leave-one-class-out and prior sensitivity tests.
• Separate a permitted population from evidence that any object is primordial.

Source and data map

Carr et al. 2026 provide the current broad synthesis and link the PBHconstraints repository, which contains digitized tables and a Mathematica notebook. Carr et al. 2017 supply the conditional extended-function method; Escrivà et al. supply formation physics; Auffinger reviews evaporation and messenger dependencies 2023, §§2–9; Andrés-Carcasona et al. provide the current O4a solar-mass row; Carr et al. 2021 and Franciolini et al. supply independent critical syntheses.

No end-to-end rerun was performed. The PBHconstraints notebook is a reproduction aid, not a substitute for each raw likelihood. Its commit was not authoritatively tagged to the 2026 journal article, and every digitized input was not traced to raw products. No durable code and likelihood archive was verified for arXiv:2605.15749v2. A reproduction must pin the population, source likelihoods, nuisance priors, covariance, response linearity, confidence semantics, code commits, and file checksums.

References

• Andrés-Carcasona, M., Iovino, A. J., Vallejo-Pagès, E., Vaskonen, V., Veermäe, H., Martínez, M., and Mir, Ll. M. 2026, “Constraints on Primordial Black Holes from the First Part of LIGO–Virgo–KAGRA Fourth Observing Run,” arXiv:2605.15749v2, preprint, arXiv.
• Auffinger, J. 2023, “Primordial Black Hole Constraints with Hawking Radiation—A Review,” Progress in Particle and Nuclear Physics 131, 104040, DOI 10.1016/j.ppnp.2023.104040.
• Carr, B., Raidal, M., Tenkanen, T., Vaskonen, V., and Veermäe, H. 2017, “Primordial Black Hole Constraints for Extended Mass Functions,” Physical Review D 96, 023514, arXiv:1705.05567v3, DOI 10.1103/PhysRevD.96.023514. Open PDF.
• Carr, B., Kohri, K., Sendouda, Y., and Yokoyama, J. 2021, “Constraints on Primordial Black Holes,” Reports on Progress in Physics 84, 116902, DOI 10.1088/1361-6633/ac1e31.
• Carr, B., Iovino, A. J., Perna, G., Vaskonen, V., and Veermäe, H. 2026, “Primordial Black Holes: Constraints, Potential Evidence and Prospects,” Rivista del Nuovo Cimento 49, 225–274, arXiv:2601.06024v2, DOI 10.1007/s40766-026-00080-z.
• Escrivà, A., Kühnel, F., and Tada, Y. 2022, “Primordial Black Holes,” arXiv:2211.05767v5, arXiv.
• Franciolini, G., et al. 2024, “Primordial Black Holes and Their Gravitational-Wave Signatures,” Living Reviews in Relativity 27, 1, DOI 10.1007/s41114-024-00053-w.