Camshaft Selector — What Cam Do I Need?

"What cam do I need?" is the most-asked question in engine building — and the most-botched. The answer isn't a part number, it's a set of specs (duration at 0.050", lobe separation angle, and lift) that match your displacement, compression, transmission, and goal. Pick the specs above and this tool recommends a cam and shows the powerband it makes, how it idles, and the converter, gear, and compression it needs to actually work. The LSA math follows David Vizard's published method; the duration-to-RPM mapping follows dyno-verified manufacturer data.

The three numbers that define a camshaft

Every cam recommendation comes down to three specs. Get these right and the part number sorts itself out.

• Duration at 0.050" — how long the valve is meaningfully open, measured at 0.050" of lifter rise. This is the only cross-comparable duration number (every manufacturer measures it the same way), so it's what you shop by. More duration moves the powerband up in RPM at the cost of low-end torque and idle quality.
• Lobe Separation Angle (LSA) — the angle in cam degrees between the intake and exhaust lobe peaks. It sets valve overlap, which controls idle quality, vacuum, and how wide the powerband is. Tighter LSA = more overlap = lopier idle + more peak torque; wider LSA = smoother idle + broader band.
• Valve lift — how far the valve opens. More lift helps only if the cylinder head can flow the extra — past a head's useful lift, you're just stressing the valvetrain. Most OE valvetrains top out around 0.550".

Advertised duration (the bigger number on the cam card) is measured at different checking heights by different makers (COMP at 0.006", others at 0.004" or 0.007"), so two cams with the same advertised duration can be very different. Only use it to cross-check — always compare the 0.050" number.

Duration @ 0.050" by application

These ranges are for an SBC-class V8 (~350ci). Bigger engines can run a bit more duration for the same drivability because they still make idle torque; smaller engines want a little less. The calculator scales the recommendation to your displacement automatically. LS engines are the exception: makers grind LS cams smaller for the same use (towing about 206–209°, street 212–226°, street/strip around 227°), so the calculator follows the makers' own LS cams there.

Rule of thumb: every +10° of duration at 0.050" moves the powerband up roughly 500 RPM (about 50 RPM per degree). Dyno testing of COMP's XR-HR hydraulic rollers on a 460 big-block confirmed the shape — going from 224° to 236° at 0.050" pushed the HP peak from 5,500 to 5,800 RPM (Street Muscle, "The Duration Game").

How LSA is chosen — Vizard's method

Lobe separation angle isn't guesswork. David Vizard's widely-used method sets the ideal LSA from cylinder displacement divided by intake valve diameter — valve size stands in for how well the head flows. His "128 rule" for parallel-valve V8s:

LSA = 128 − ( (CID ÷ cylinders ÷ intake-valve-diameter) × 0.91 )

The 128 constant is Vizard's for the small-block Chevy and other parallel-valve V8s. For the big-block Chevy we use 131, which lands within about a degree of the LSAs in Vizard's big-block Chevy book (107–108° for a 468, 104° for a 582 on 2.3" valves). The constants for the other head families (127 Ford Windsor, 129 Pontiac, 130 Mopar big block, 131 Cleveland and 385-series) are our own estimates; we haven't found them in a book or magazine yet. Then Vizard adds +0.75° of LSA for every point of compression over 10.5:1 (and correspondingly less below), because the base formula assumes about a 10.5:1 engine.

Worked example — 350 SBC, 2.02" intake valve, 10.0:1:

• 350 ÷ 8 ÷ 2.02 = 21.66 cubes per inch of valve
• 128 − (21.66 × 0.91) = 128 − 19.71 = 108.3° base LSA
• Compression is under 10.5, so essentially no adjustment → ~108° LSA

A 383 with the same heads works out to ~106° (more cubes per inch of valve = tighter LSA). These match the values our Advanced Cam Calculator produces. Vizard's own caveat: the real relationship is a gentle curve, not a perfectly straight line, so treat the output as a strong starting point, not gospel.

How the calculator sets LSA. Vizard's rule gives the tightest LSA an engine makes the most power on. Cam makers grind wider for street manners: across 3,200+ catalog cams the median is about 112° on mild cams, 110° on street cams and 106–108° on big race cams, and LS cams sit at 112–114° almost across the board. Taken alone, the rule came out about 3° tighter than the makers' cams for the same use. So the calculator starts from the makers' median LSA for your engine family and cam size, then uses Vizard's rule to adjust it: tighter for more cubic inches per inch of valve than the family's typical engine, wider for more compression, and wider again for boost or nitrous. Checked against 77 hand-verified maker cams with a stated use, it lands within about 1° of the maker's LSA on the median.

What LSA does to idle, vacuum, and powerband

LSA controls overlap — the moment when both valves are open near TDC. More overlap (tighter LSA) is what makes a cam sound lopey, and it costs idle vacuum. A CarTech dyno test of five cams (identical except LSA) on one engine measured idle vacuum at 1,000 RPM. They were solid rollers of about 235/240° at .050", so a milder hydraulic street cam pulls more vacuum than this at the same LSA:

Source: CarTech, "Camshafts Demystified". Peak HP RPM barely moved (6,200–6,400) across all five — LSA moves the torque peak and idle, not the HP peak. A power-brake car generally wants ≥ ~14 inHg at idle, which lines up with 112°+ LSA. Values scale with the engine, so treat the numbers as a representative shape (about +1 inHg per +2° LSA in the street range), not universal constants.

Vizard's overlap targets by use

Vizard actually treats overlap — not duration or LSA directly — as the primary design variable. You pick a target overlap for your application, then duration and LSA are chosen to hit it. His published overlap chart (seat-to-seat, in degrees):

The relationship: Overlap = (intake duration + exhaust duration) ÷ 2 − (2 × LSA) using advertised (seat) durations. Overlap increases about 2° for every 1° you tighten the LSA. Multiple sources corroborate this chart (Engine Builder, Speed-Talk, JEGS).

Intake valve closing — the master variable

Of all four valve events, the intake valve closing (IVC) point matters most, because it defines how much charge is actually trapped and compressed — it sets the dynamic compression ratio and therefore the low-end torque and detonation behavior.

• Early IVC (~30–50° ABDC): high dynamic compression, strong low-to-mid torque, broad powerband, higher cranking pressure. This is where street and towing cams live.
• Late IVC (~65–80° ABDC): bleeds cylinder pressure back into the intake at low RPM (killing low-end torque and dropping dynamic compression) but lets the engine keep filling at high RPM. Race territory.
• Past ~75° ABDC: most of the engine's low-speed torque is gone (Engine Builder, "Valve Timing Events and the Order of Importance").

This is why a big cam feels lazy in a low-compression engine: the late intake close drops dynamic compression, and there's not enough static compression to make up for it. The fix is more static compression — which is why the calculator flags a compression bump when you pick a big cam for a low-compression engine. By our octane estimate, 93 octane allows about 8.45:1 dynamic compression (roughly 175 psi cranking) with iron heads and about 9.7:1 (roughly 210 psi) with aluminum heads. For the low end, Vizard wants at least 180 psi for a modest street engine, and 200, 220 and 240 psi for performance street, street/strip and race (Vizard via ChevyDIY). Check yours with our dynamic compression calculator.

The #1 cam-selection mistake: buying too much cam

The single most common error is choosing a cam bigger than the rest of the combination can support. A cam only works inside its powerband — if the converter, gears, and compression don't put the engine into that band, the car is slower than it would be with a smaller cam.

The whole combo has to agree:

• Converter stall (automatics): the stall speed should put the engine into the cam's powerband off the line. A cam that comes alive at 3,000 RPM behind a stock 1,800-RPM converter feels flat and bogs. Rule: stall near where the cam "comes on."
• Rear gear: more cam wants more gear to keep the engine spinning in its band. Street cams like 3.42–3.73; strip cams want 3.73–4.10+.
• Compression: a bigger cam bleeds cylinder pressure — raise static compression to compensate, or the engine feels soft and lazy down low.
• Cylinder heads: lift past the head's useful flow does nothing. A big cam on stock heads is wasted; matched heads/cam/intake (an "HCI" package) is how you actually make power.
• Valve springs: every performance cam needs springs matched to its lift and RPM. Stock springs float at high RPM and can drop a valve.

The calculator flags these based on your inputs. The lesson from decades of dyno testing: a right-sized cam in a matched combo beats a big cam in a mismatched one, every time.

Boost and nitrous change the rules

Forced induction and nitrous want wider LSA than an equivalent naturally-aspirated cam, because the power adder does the cylinder-filling work — you need less overlap, and wide LSA reduces reversion against boost backpressure.

• Supercharged / Turbocharged: add roughly 4° of LSA vs the NA number. Turbos especially benefit from earlier exhaust closing to cut reversion from exhaust backpressure during overlap (OnAllCylinders supercharged cam test).
• Nitrous: add about 2° of LSA and often more exhaust duration. Too much overlap sends nitrous straight out the tailpipe unburned (LSXMag nitrous cam test).

The calculator applies these LSA adjustments automatically when you set the aspiration.

Advanced mode — replace the assumptions with real numbers

Simple mode makes reasonable assumptions to get you a fast answer: it uses a standard rocker ratio for lift, a typical bore and stroke for the displacement, and a dual-plane intake. Switch to Advanced mode to override them with data you actually have:

• Intake manifold type: a dual-plane favors low and mid-range and signs off up top; a single-plane shifts the whole band up. The calculator moves the estimated powerband accordingly.
• Target peak-HP RPM — if you know the RPM you want peak power at (say you're building around a converter and gear), the calculator reverse-solves the duration to land there instead of using the application preset.
• Stroke + rod length — combined with your static compression, these let the calculator compute the dynamic compression ratio and cranking pressure the recommended cam will actually produce, and tell you whether that's pump-gas friendly. This is the honest version of the "does a big cam need more compression?" question — it uses the recommended cam's real intake-closing point.

The dynamic-compression math uses the standard slider-crank effective-stroke method: DCR = 1 + (effective stroke ÷ full stroke) × (static CR − 1), where the effective stroke is measured from the intake-valve-closing point. Because bore and chamber volume cancel out of that ratio, we only need stroke, rod, static compression, and the cam's IVC — no chamber cc required. Cranking pressure follows the polytropic model (14.7 × DCR^1.2 − 14.7). The calculator then gives the octane that dynamic compression wants, with iron and with aluminum heads, from the same estimate the compression calculator and the Build Advisor use.

Any advanced field you leave blank simply falls back to the simple-mode estimate, so you can fill in only the numbers you have.

Rocker ratio and lift. A cam's lobe lifts the lifter; the rocker multiplies that into valve lift. Lobes are much alike across engine families (about .35" for a 225° hydraulic roller), so the valve lift makers print depends mostly on the family's stock rocker: 1.5 for small-block Chevy, Pontiac and Mopar LA/B/RB, 1.6 for small-block Ford and Magnum, 1.7 for LS and big-block Chevy, 1.73 for Cleveland and 429/460, 1.8 for LS7 and Gen V LT. The calculator works lift out that way, and shows the matching cams' lift at your ratio if you enter one. A higher-ratio rocker on the same cam is worth about 0–3% per 0.1, mostly at the top end, and only while the heads still gain flow at the extra lift; it also raises valve acceleration, so check springs and clearances.

Compression window. For each recommendation the calculator gives the minimum static compression cam makers print for a cam that size (COMP, Crane, Lunati, Crower, Isky, Erson and Elgin agree within about half a point: about 9.0:1 for mild cams, 9.5 in the 220s, 10.0 in the 230s, 10.5 in the 240s and 11.0 past 250°) and the most 93 pump gas takes with iron and with aluminum heads, both from the cam's intake closing point. A bigger cam closes the intake later, so it needs more static compression to feel crisp and tolerates more before it knocks.

Heads first, then the cam

The same cam peaks at a very different rpm depending on the heads. On 123 documented pushrod dyno builds, a rule that predicts peak power from duration alone read about 1,100 RPM low, because most of those builds ran good aftermarket heads. Better heads let the engine keep filling at higher rpm, so peak power moves up and grows. A cam bigger than the heads can feed only moves torque higher and costs low end.

So the calculator lets you say which heads you have (optional): stock, street aftermarket, ported or performance, race, or your own flow number. Left unset, it assumes typical factory heads for your bore. The powerband, peak rpm and power all come from the Build Advisor's power model, which was fitted to those dyno builds. The results include a same cam, different heads table for your engine family, so you can see how much the heads alone change. The typical flows are from factory-head flow tests (stock) and published flow sheets for the family (the rest).

DOHC and 4-valve engines follow different rules (cams rated at 1 mm, power set around the factory cam and redline), so for those the calculator points you to the Build Advisor.

Shopping the recommendation — real matching cams

Below the recommendation, the calculator lists real, currently-available grinds from the major manufacturers whose specs land closest to your target duration, LSA, and lift. The database covers popular street and street/strip cams from COMP Cams, Lunati, Crane, Crower, Howards, Edelbrock, Summit Racing, plus LS-specific makers like Brian Tooley Racing, Texas Speed, Cam Motion, and Trick Flow — across Small Block Chevy, LS/LT, Big Block Chevy, Ford Windsor, Ford 351 Cleveland, Ford 429/460, Mopar small block (LA), Mopar big block (B/RB), and Pontiac V8.

Every listed cam's duration at 0.050", LSA, lift, and part number was pulled from a published manufacturer or retailer spec page. But always verify before you buy — specs and part numbers change, listed lift assumes the platform's standard rocker ratio (so your actual lift may differ), and you must confirm fitment and piston-to-valve clearance for your specific combination. The matches are shopping starting points, not endorsements, and engine-build.com is not affiliated with any cam manufacturer.

Frequently asked questions

What cam do I need for a stock 350?
For a stock or mildly-modified 350 you want a mild-street grind: roughly 218–224° duration at 0.050", around 110–112° LSA, and 0.500"-ish lift. This keeps a slight lope, works with a stock or low-stall converter, and doesn't require compression or head changes. A classic example is a 224°/230° hydraulic roller on a 110–112° LSA.

How do I choose duration at 0.050"?
Match it to your target RPM band. Street engines live in the 214–232° range; street/strip in the 232–244° range; race above that. Every +10° moves the whole powerband up about 500 RPM. Pick the duration whose powerband centers on where you actually drive or race.

What does LSA (lobe separation angle) do?
It sets valve overlap, which controls idle character and powerband width. Tighter LSA (106–110°) = lopier idle, more peak torque, narrower band. Wider LSA (112–116°) = smoother idle, more vacuum, broader band. It moves the torque peak and idle, but barely affects where peak horsepower lands.

Should I run a tight or wide LSA?
Tight (108–110°) for a track-focused or torque-monster build that doesn't need a smooth idle or power brakes. Wide (112–114°) for a street car that needs vacuum, a smooth-ish idle, and a broad, forgiving powerband — and always wider for boost or nitrous.

Does a bigger cam need more compression?
Yes. A longer-duration cam closes the intake valve later, which bleeds off cylinder pressure at low RPM and drops dynamic compression. Raising static compression restores the lost cylinder pressure and keeps the engine responsive. As a rule, step compression up as you step cam duration up — cam makers ask about 10:1 for a cam in the 230s at .050", not 9:1.

What converter stall do I need for my cam?
The stall speed should put the engine into the cam's powerband off the line — roughly where the cam "comes alive." A mild-street 220° cam is happy with ~2,000–2,400 RPM; a 236° street/strip cam wants 3,000+ RPM. Too tight and the car bogs; the calculator estimates the range for your cam.

How much lift is too much?
Lift past the cylinder head's useful flow range does nothing but stress the valvetrain. Most OE valvetrains (springs, retainers, guides) top out around 0.550" lift. Above that you need upgraded springs and a piston-to-valve clearance check. More lift only helps if the head can flow the extra air.

Hydraulic roller, solid roller, or flat tappet?
Hydraulic roller is the modern default: no lash adjustment, good RPM capability, less wear than flat tappet, and works on the street. Flat tappet is cheaper but needs careful break-in and is being phased out. Solid roller is for serious race engines that spin past ~6,500 RPM and don't mind periodic lash checks.

What's the difference between advertised duration and duration at 0.050"?
Advertised duration is measured at a small checking height (often 0.006"), so it captures the slow opening ramps; duration at 0.050" is measured after the valve is meaningfully open. The gap is typically 45–55°, but because each manufacturer advertises at a different height, advertised numbers aren't cross-comparable. Always shop by the 0.050" number.

Will this cam idle rough?
It depends on overlap, which the calculator estimates. Under ~35° overlap is a smooth idle; 50–65° is a noticeable healthy lope; 80°+ is a rough race idle that needs a high idle speed and won't pull much vacuum. The idle-vacuum estimate tells you whether power brakes will still work.

Where this method comes from

The LSA rule, the compression adjustment and the flow-to-horsepower figure used above are David Vizard's, published in David Vizard's How to Build Horsepower (CarTech, ISBN 9781934709177). His How to Port & Flow Test Cylinder Heads covers the flow bench work these figures rest on, and How to Build Max-Performance Chevy Small-Blocks on a Budget applies the same approach to one platform in detail.

This calculator implements part of that method. The books set out the reasoning, the dyno work behind the numbers and — more usefully — where the method stops applying. If you are about to spend four figures on parts, they are worth reading rather than trusting a web form.