A decade-long experiment at the National Institute of Standards and Technology has yielded a new measurement of the universal gravitational constant, or G, that diverges unexpectedly from another leading result. The mismatch is minuscule in absolute terms, but for one of nature's most fundamental quantities it is large enough to sustain a puzzle that has persisted for 225 years.
The finding, reported after the experiment's sealed data were finally opened, adds fresh tension to a long-running problem in physics: G is the least precisely known of the fundamental constants, and independent measurements of it have stubbornly refused to converge. The new NIST value differs from another prominent result by an amount that exceeds what the stated uncertainties would predict, a pattern that has now survived yet another high-precision attempt.
Researchers involved in the work say the discrepancy could point to unrecognized experimental errors in one or more of the measurements. Alternatively, it could hint at something more surprising about the nature of gravity itself. The experiment was designed to reduce known sources of error, and its results were kept sealed for ten years before being analyzed, a safeguard intended to prevent unconscious bias from shaping the outcome.
G describes the strength of the gravitational attraction between masses. Unlike constants such as the speed of light, which can be measured with extraordinary precision, G is extraordinarily weak and cannot be isolated from the gravitational pull of surrounding objects. Every laboratory measurement must account for the mass of instruments, walls, floors, and distant terrain, making the task a test of experimental discipline as much as of theory.
The constant's history reflects that difficulty. The first laboratory measurement of G was performed by Henry Cavendish in 1798, and in the more than two centuries since, successive experiments have produced values that scatter more widely than their claimed uncertainties suggest. The new NIST result joins that lineage, reinforcing rather than resolving the spread.
For now, the discrepancy remains small enough that it does not threaten the use of G in most calculations. But it matters for precision tests of gravity, for efforts to detect deviations from established theory, and for the broader question of whether the laws of physics are as uniform as they appear. If the scatter is caused by overlooked systematic effects, the fix lies in better instrumentation and methodology. If it is not, the implications could reach further.
The NIST team's approach involved years of careful measurement and a deliberate delay before examining the data, a method meant to keep expectations from influencing interpretation. That the result still conflicts with another leading value suggests the problem is not easily dismissed as a single flawed experiment.
Physicists working on the constant say the path forward lies in new measurements using independent methods, ideally ones that share as few sources of error as possible. Until those results arrive, G will remain a rare case in modern physics: a fundamental quantity whose value is known well enough for everyday science but not well enough to satisfy the field that measures it.
The puzzle is not new, but each new high-precision result sharpens it. The NIST measurement, by failing to align with a competing value, keeps open the possibility that the discrepancy is telling scientists something they have not yet learned how to hear.
