In 1928, a highway engineer in California had a problem nobody had solved properly. Roads were cracking apart under trucks that hadn't existed a decade earlier. Nobody had a reliable way to say, before construction started, whether the ground underneath a road could actually take the load. His name was O.J. "Pappy" Porter. He worked for the California Division of Highways (now called Caltrans) and instead of guessing, he built a test. Ninety years on, engineers still call it by his employer's name: the California Bearing Ratio test.
Key Takeaways
- CBR is a percentage: how a soil's resistance to a standard plunger compares to perfect crushed stone, tested per IS 2720 (Part 16): 1987.
- Soaked CBR, which submerges the sample for 96 hours before testing, is the standard for Indian design. An unsoaked value on a site that will meet monsoon water is measuring the wrong condition.
- IRC 37 sets a minimum design subgrade CBR of 5%. Below that, the standard response is soil stabilisation, not a thicker pavement.
Push a small steel plunger into soil at a fixed speed. Measure how hard the ground pushes back. Compare that to how hard a slab of perfectly crushed stone pushes back under the same conditions. That's the whole idea. California adopted it officially in 1935. It's now written into ASTM, AASHTO, British Standards, and, most relevant here, IS 2720 (Part 16): 1987, reaffirmed as recently as December 2021.
Nearly a hundred years later, almost every kilometre of highway, every airport runway, every industrial yard in this country still gets built on the back of that one number. It's called the California Bearing Ratio, or CBR. If you're reading this, you're probably about to make a decision that depends on it.
What Does CBR Actually Measure?
Every road is really a sandwich. Layers of increasingly strong material sit on top of whatever soil happens to already be there: the subgrade. You don't choose the subgrade; it's the site you were handed. What you do choose is how thick and how strong the layers above it need to be to stop that subgrade from failing under load. Weak subgrade means a thicker, more expensive pavement. Strong subgrade means you can often build lighter. CBR is the number that tells you which situation you're actually in, before construction starts.
The formal definition: CBR is the ratio, expressed as a percentage, of the force needed to push a standard 50 mm diameter plunger into a soil sample at 1.25 mm per minute, compared to the force needed for the same penetration into standard crushed stone. The crushed stone is arbitrarily assigned a CBR of 100%. A CBR of 8% means your soil offers 8% of that resistance, not a failing grade, just a normal reading for a lot of Indian subgrade soil, and the exact number a pavement designer needs to compensate for.
What Happens Inside a CBR Test?
Per IS 2720 (Part 16), every dimension in this test is fixed, so a result from one lab is directly comparable to a result from any other lab in the country. The mould is 150 mm internal diameter and 175 mm high. A 60 mm extension collar sits above it during compaction. A 148 mm spacer disc occupies the space the base layer would take up. The compaction rammer weighs 4.89 kg and drops through 450 mm. Surcharge weights of 2.5 kg each simulate the pavement layers that will eventually sit above the soil. The loading machine holds a controlled penetration rate of exactly 1.25 mm per minute. None of these numbers are arbitrary — change the rammer weight and you change the compactive effort, and you get a different, wrong, CBR.
Soaked or Unsoaked — Why This Decision Matters More Than Almost Anything Else
Before compaction even starts, one decision changes the entire outcome. Unsoaked CBR measures the soil as it is right now. Soaked CBR submerges the compacted mould in water for 96 hours first, forcing the soil to absorb as much moisture as it will realistically see over its service life. For anything built in India, soaked CBR is the standard the design should be based on. This is the single most common place non-specialist testing goes wrong: an unsoaked value looks reassuringly high, gets used for design, and the first real monsoon exposes a subgrade that was never actually tested under the conditions it would face.
The Penetration and the Number It Produces
The plunger drives into the soil at 1.25 mm per minute. Load readings are recorded at fixed intervals, from 0.5 mm out to 12.5 mm, using a calibrated proving ring and dial gauge. Two points on the resulting curve matter most: 2.5 mm and 5.0 mm penetration. Each is compared against the standard crushed-stone loads: 1,370 kgf (13.44 kN) at 2.5 mm and 2,055 kgf (20.15 kN) at 5.0 mm. Normally the 2.5 mm value is reported. If the 5.0 mm value comes out consistently higher on a repeat test, that value is used instead — a built-in check against a sample that behaved unusually in the first few millimetres.
What Does Your CBR Value Actually Mean for the Project?
A CBR percentage on its own is just a number. What makes it useful is what Indian pavement design does with it, governed by IRC 37, the Indian Roads Congress guideline for flexible pavement design. The lower the CBR, the thicker and more expensive the pavement structure has to be to compensate. IRC 37 sets a minimum design CBR of 5% for subgrade; below that, the response isn't a thicker road, it's improving the subgrade itself, through replacement, mechanical stabilisation, or lime/cement treatment.
| CBR Value | What It Generally Means |
|---|---|
| Below 2% | Subgrade essentially unsuitable as-is; stabilisation usually mandatory before design proceeds |
| 2% – 5% | Poor subgrade; pavement design carries a real cost penalty in extra thickness |
| 5% – 10% | Workable, standard subgrade, where most real Indian sites land |
| Above 10% | Strong subgrade; layer thicknesses can often be optimised down |
Two sites four points apart on CBR can end up with pavement designs that differ by tens of lakhs in material cost across a few kilometres. That's the entire reason this test exists.
How Is CBR Calculated? A Worked Example
Say a soaked test on a given soil records a load of 550 kgf at 2.5 mm penetration:
CBR = (550 ÷ 1370) × 100 ≈ 40%
A CBR of 40% is a strong subgrade by Indian standards — well above the IRC 37 minimum, the kind of number that lets a designer trim layer thickness rather than pad it out. This example is illustrative only; every real result depends on the specific soil, moisture history, and compaction achieved on your site, which is exactly why the test has to be run on your actual material.
Where Does CBR Testing Go Wrong?
Compacting at the wrong moisture content. CBR is run at the soil's Optimum Moisture Content and Maximum Dry Density, determined separately via a Proctor test. Skip that step and the result measures a soil condition that never actually existed on site.
Treating unsoaked results as design-ready on a site that will meet monsoon water. Disturbed or non-representative sampling. A CBR test is only as good as the sample that went into the mould. Using field CBR and lab CBR interchangeably. Field or in-situ CBR (estimated through methods like Dynamic Cone Penetrometer correlation, or Global Lab's own field CBR testing) is genuinely useful for rapid screening across a large site or verifying compaction consistency. It is not a substitute for the controlled, soaked laboratory procedure when it comes to actual pavement design. Both have a place; they're not the same test.
What Other Tests Does CBR Need to Make Sense?
CBR rarely stands alone in a real geotechnical report. It's usually read alongside Atterberg limits, which classify how the soil behaves as moisture changes; specific gravity and water content, which feed directly into the compaction calculations CBR depends on; particle size distribution, which explains why one soil compacts cleanly and another doesn't; and free swell index, which flags expansive soils that can undermine a pavement design no CBR number alone would catch. A plate load test often follows at the field stage to confirm what the lab CBR predicted. None of these replace CBR — they're what turns a single percentage into a design you can actually build on.
Frequently Asked Questions
A pavement doesn't fail on day one. It fails eighteen months to three years in, quietly, and tracing that failure back almost always leads to a CBR value that was skipped, estimated, or run without soaking on a soil that was always going to meet water. Global Lab runs CBR testing under IS 2720 (Part 16) as an NABL-accredited laboratory, with both soaked and unsoaked protocols. Get in touch or explore our full field testing of soil services for the complete geotechnical picture.
Sources:
IS 2720 (Part 16): 1987, Reaffirmed 2021, Bureau of Indian Standards — Methods of Test for Soils: Laboratory Determination of CBR
IRC 37 — Guidelines for the Design of Flexible Pavements, Indian Roads Congress
Indian Railway Institute of Civil Engineering — California Bearing Ratio Test (accessed 15 September 2026)
Further reading on CBR testing per IS 2720, from our sister lab: JS Civil Lab Solutions (accessed 15 September 2026)