Abstract Assembly fixtures are commonly used to locate rigid and compliant components for joining operations. Piezoelectric (PZT) impedance sensors can be used to individually monitor assembly fixtures and have been proven effective for structural health monitoring applications. The sensors rapidly record significant quantities of data, but must be bonded to a fixture in order to transmit an input signal and record the corresponding output signal. Thus, PZT impedance sensors become a permanent feature of an assembly fixture and may create unique systems defined by the assembly fixture and impedance sensor (AFIS). Previous research has shown success in detecting fixture damage using PZT impedance sensors. This paper extends previous fixture damage detection work to damage diagnosis through the use of data mining classifiers. Classifiers were used in three studies; the first was to show that classifiers can be trained to classify a healthy fixture, fixture damage, and multiple severities of fixture damage in an isolated AFIS. In the second study, classifier generalization was tested by simulating an unknown damage. Lastly, classifiers were used to study the uniqueness (i.e. fixture classification) of two AFIS, which could have implications related to the practical application of classifier models to any AFIS.