Short answer. A turning movement count records every vehicle passing through an intersection by its approach and its movement — left, through or right — usually in 15-minute intervals. Those intervals are used to identify the true peak hour, calculate a peak hour factor that describes how evenly demand arrives within it, and provide the movement-level volumes that capacity analysis, signal timing and traffic impact studies require.
What the count records
At a four-leg intersection, each approach has three basic movements — left, through and right — giving twelve vehicle movements in total, plus pedestrian and cyclist movements on each crosswalk. A turning movement count records each of those separately over a defined period. It is not a volume count with extra steps; the movement-level split is what makes it useful, because capacity, delay and queueing all operate at the movement level, not the intersection level.
Movements are named by the direction of the approach, not the direction of travel after the turn. A vehicle travelling north that turns left is a northbound left, even though it ends up heading west. Getting this convention right matters when the counts feed a model, because a transposed naming scheme produces a plausible-looking model of a different intersection.
Why 15-minute bins
Counts are tallied in 15-minute intervals for two reasons. First, the peak hour is discovered, not assumed: the highest four consecutive 15-minute bins define it, and that hour frequently does not align with clock hours. A peak that runs 7:45 to 8:45 will be understated by anyone who counted 8:00 to 9:00.
Second, the sub-hourly variation within the peak hour is itself a required input. That is the peak hour factor.
Peak hour factor, and why the default is dangerous
The peak hour factor is the peak-hour volume divided by four times the highest 15-minute volume within that hour:
PHF = V(hour) ÷ (4 × V(peak 15 min))
A PHF of 1.0 means demand arrived perfectly evenly across the hour. A PHF of 0.75 means it was heavily concentrated in one 15-minute surge, and the intersection has to accommodate that surge, not the hourly average.
Typical values run from about 0.85 to 0.95 for urban intersections and lower — sometimes 0.70 to 0.80 — at rural sites, at school-adjacent locations, or anywhere a single large generator releases traffic in a burst. Capacity analysis uses PHF to convert an hourly volume into the flow rate the intersection must actually serve, so it directly scales the demand figure.
This is why a defaulted PHF is a review risk. Carrying 0.92 through the analysis when the observed value is 0.78 understates the design flow rate by roughly 18%, which can move a movement from acceptable to failing. The 15-minute data exists precisely so that this figure is measured. Use it.
Vehicle classification
How finely vehicles are classified should be decided by what the analysis needs, because classification detail is the main driver of processing cost.
| Scheme | Categories | Use case |
|---|---|---|
| Basic | Light / heavy | Simple volume reporting |
| Standard 3-class | Car, single-unit truck, articulated | Most capacity analysis and TIA work |
| Extended | Adds bus, motorcycle, bicycle, light commercial | Multimodal and transit corridors |
| FHWA 13-bin | Full federal classification | Pavement design, federal reporting, freight studies |
| Local mixed traffic | Adds auto-rickshaw, two-wheeler categories, non-motorised | South Asian and mixed-traffic contexts |
Heavy-vehicle percentage by movement matters more than the total, because a 12% heavy share concentrated in one left-turn lane affects that lane's saturation flow far more than the same share spread evenly across the intersection.
How to specify a count
- Location and approach naming, with a site map or aerial showing how each leg is labelled
- Date and day of week, avoiding holidays, school breaks, adjacent construction and unusual weather
- Survey periods, for example 07:00 to 09:00 and 16:00 to 18:00, or a 12-hour 07:00 to 19:00 count
- Interval length — 15 minutes is standard; 5-minute bins where surge behaviour is being studied
- Classification scheme, named explicitly rather than described as detailed
- Whether pedestrians and cyclists are included, and by which crosswalk legs
- Whether signal phase or cycle logging is required
- Deliverable format — your Excel template, and whether movement diagrams are needed
Errors that invalidate a count
- Counting a clock hour instead of finding the true peak hour from the interval data
- Applying a default peak hour factor when 15-minute data was collected
- Camera position that occludes one approach, producing a systematic undercount rather than random error
- Counting during an atypical week — a nearby closure, an event, a school holiday
- Mislabelled approaches, which produce a coherent model of a different intersection
- Classification inconsistency across a long shift, particularly light commercial versus car
- Missing the shoulders, so a peak that starts before the survey period is truncated
Reading a count sheet
A well-formed count workbook has an interval sheet with every 15-minute bin for every movement, a peak-hour summary showing which four bins were selected and the resulting PHF per approach, a classification breakdown, and a diagram. If a count arrives as hourly totals only, the underlying interval data has been discarded and you cannot verify the peak hour or recalculate the PHF — which is a good reason to insist on interval data in the specification.
Frequently asked questions
What does a turning movement count measure?
Every vehicle through an intersection, categorised by the approach it arrived on and the movement it made — left, through or right — normally recorded in 15-minute intervals, together with pedestrian and cyclist crossings where specified.
How do you calculate peak hour factor?
Divide the peak-hour volume by four times the highest 15-minute volume inside that hour. A result near 1.0 means even arrivals; a lower value means demand was concentrated in a short surge.
How long should a turning movement count run?
AM and PM peak periods of two to three hours each is the common standard, since this captures the peak plus its shoulders. Twelve-hour counts are used where an agency requires daily profile data.
What is a good peak hour factor?
There is no good value — it is a site characteristic, not a quality measure. Urban intersections typically fall between 0.85 and 0.95; sites near a single large generator can be much lower, and that lower figure is the one your analysis needs.
Should pedestrians be counted with vehicles?
Almost always yes. Adding pedestrian and cyclist counts to the same footage costs very little compared with a separate survey, and most agencies now expect multimodal data in a submission.
The service behind this guide
Turning Movement Counts
Left, through and right volumes for every approach, binned to 15 minutes, peak-hour summarised and formatted to your template.
See the service → MethodsTraffic Data Collection Methods Compared
Seven ways to collect traffic data, what each can actually measure, and why the cheapest method per hour is often the wrong one for the question you are asking.
12 min read → Traffic impact studiesTraffic Impact Study: A Complete Guide for 2026
A traffic impact study is the analysis a jurisdiction requires before approving a development that will add trips to the network. Here is what one contains, when you need it, and what it rests on.
11 min read →Need this data collected?
We process traffic, parking and pedestrian surveys from your video and return them in your own template, typically within 48 hours.