Car Diagnosis is an example of Bayesian Network structure originally designed by Norsys Software Corp. The example structure shows the relationships between different parts of a car. This network includes 15 discrete variables. Label values of the variables are annotated in the following table.
| Variables | Label 1 | Label 2 | Label 3 |
| Voltage_at_plug | 1=none | 2=weak | 3=strong |
| Car_cranks | 1=False | 2=True |
| Spark_plugs | 1=fouled | 2=too wide | 3=okay |
| Spark_quality | 1=very bad | 2=bad | 3=good |
| Distributer | 1=faulty | 2=okay |
| Spark_timing | 1=very bad | 2=bad | 3=good |
| Car_starts | 1=False | 2=True |
| Headlights | 1=off | 2=dim | 3=bright |
| Alternator | 1=faulty | 2=okay |
| Charging_system | 1=faulty | 2=okay |
| Battery_voltage | 1=dead | 2=weak | 3=strong |
| Battery_age | 1=very old | 2=old | 3=new |
| Main_fuse | 1=blown | 2=okay |
| Starter_system | 1=faulty | 2=okay |
| Starter_motor | 1=faulty | 2=okay |