If the starter turns the engine but it will not run, separate usable cranking power, engine-position information, fuel delivery, ignition and mechanical condition. A pump noise, fuel smell or missing code does not establish which part failed. Capture evidence during the actual no-start before the engine starts working again.
This guide concerns a conventional gasoline engine that actually rotates during a start attempt. A click with no engine rotation is a different complaint. Diesel, hybrid high-voltage and automatic stop-start diagnosis require their own procedures. Use the vehicle's starting instructions and cranking limits; do not keep trying until the battery is flat.
First describe what is turning
Write down whether the engine cranks slowly, at its usual rhythm, unusually fast or unevenly. A starter motor can spin without turning the engine, so a whirring sound alone is not confirmation. Unusual mechanical noise, smoke, a fuel leak or strong raw-fuel smell is a reason to stop, switch off and arrange assessment. Do not run starting tests in an enclosed space.
Record the warning messages, whether the engine briefly fires, and whether the failure is cold, hot or intermittent. Note recent refuelling, battery work or repairs. These observations narrow the next check without becoming a diagnosis. For example, a security message calls for the owner's manual and an authorized diagnostic procedure, not an attempt to bypass the immobilizer.
Save a short no-start record
With the vehicle safely parked and ignition in the specified mode, save the full scan report before clearing codes or disconnecting the battery. Include the engine module and any relevant start-authorization or body modules the tool genuinely supports. An engine-only scan is not a complete report from every system.
- Record current and stored code status, available freeze frame and the exact dashboard message.
- If the procedure permits a brief attempt, log supported engine RPM and supply-voltage data during that attempt, not only with the ignition on.
- Record any supported start-permission, synchronization or commanded/actual fuel-pressure information, with its units and scanner definition.
- Stop within the vehicle's cranking limit. Mark missing or unverified data as unavailable rather than entering zero.
Do not add a long PID list just because the scanner offers one: an update rate too slow for the event can hide a brief change. Keep the original freeze frame separate from your new log. A stored fault condition and the current start attempt may be different events.
Usable cranking power comes before parts guesses
A normal-looking resting battery voltage does not establish what happens during cranking. The battery, cable connections, grounds, starter and engine load interact. A starter can still rotate the engine while the electrical supply or cranking speed is unsuitable for reliable starting.
Ask for a battery and starting-circuit assessment under the actual failed condition, with cranking speed and voltage-drop results compared with the vehicle specification. Do not apply one voltage threshold to every engine or replace a starter from its sound alone. DENSO's starter troubleshooting explains why current, voltage and cranking speed must be considered together.
Does the controller see the engine turning?
A supported, updating RPM reading during cranking is useful evidence that the controller is reporting engine rotation. It does not prove correct cam/crank synchronization or every position-sensor signal. Likewise, a zero reading is inconclusive until you establish that the PID is supported, updating and being read during actual engine rotation.
If reliable data shows no rotation signal, the next step is the vehicle's signal and circuit test, not automatically a new crankshaft sensor. Wiring, power, ground, the trigger arrangement and the reporting path may need checking. DENSO describes the distinct crankshaft and camshaft inputs. Use the P0335 guide when that code is actually present, and the correlation-code guide when the evidence concerns synchronization.
Separate fuel, ignition and mechanical evidence
| Observation | Evidence still needed |
|---|---|
| The fuel pump can be heard | Actual pressure and delivery under the specified start condition, plus injector control when relevant. Sound alone does not prove fuel reaches the cylinders correctly. |
| Fuel pressure matches the specification | Whether fuel is being delivered and controlled correctly, and whether ignition, compression and timing are suitable. Pressure is one result, not a complete starting-system verdict. |
| Spark is reported during a test | Which cylinders, which test method and whether timing and the other combustion requirements are correct. One observation does not prove every cylinder fires under the required conditions. |
| Relative compression looks even | Absolute mechanical condition and timing if the complaint still points there. Similar cylinders are not necessarily healthy cylinders. |
Fuel-pressure, injector, ignition and mechanical tests need the correct equipment and vehicle procedure. Do not loosen a fuel fitting to check for delivery, apply battery power to an injector or hold an ignition lead against the engine as an improvised spark test. Direct-injection work has additional pressure, sealing and handling requirements; Bosch's high-pressure injection guidance refers technicians to its diagnostic sequence and special tools.
Compression and timing checks should follow the evidence. Relative compression is a comparison, not an absolute cylinder-pressure measurement; confirm the cylinder identification and any suspected problem with appropriate follow-up tests. Pico's compression-testing guide explains both the comparison and its limitations. These tests also require the prescribed measures to prevent unintended starting.
An example of a useful next step
Suppose the engine cranks but never fires. The scanner communicates with the ignition on, then loses communication during the start attempt. This is an invented teaching scenario, not a shop case.
The missing cranking log is not evidence that the crankshaft sensor failed. First establish whether supply voltage, module power or the communication path prevented a valid reading. Once a reliable log is available, follow the evidence it actually contains. Replacing a sensor to explain a measurement you never obtained skips that distinction.
When it starts again, preserve the comparison
Record what changed before the successful start: cooling time, an approved battery-support procedure, a repaired connection or a confirmed repair. Starting after one intervention does not, by itself, identify every cause of an intermittent complaint.
A warm restart does not verify a repair for an overnight no-start. Recheck the relevant conditions safely, retain the before/after reports and document any condition not reproduced. Use the repair-verification record to keep the symptom, finding, work and result connected. If the fault remains intermittent, a clear record of the failed attempt is useful progress even before a part is condemned.
Scope and sources
This AI-assisted educational article was prepared November 17, 2026. It is a diagnostic starting point, not a vehicle-specific repair instruction or a claim of hands-on testing. Manufacturer and equipment references support selected concepts only. Use current service information for the actual vehicle and read the editorial policy.