In this study of the project DyAdd (Adult Dyslexia and Attention Deficit Disorder in Finland), classical eyeblink conditioning (EBC) was investigated in both delay and trace paradigms in adults (18-55 years) with dyslexia (n = 37), attention deficit-hyperactivity disorder (ADHD; n = 21), their comorbid combination (n = 8), and healthy controls (n = 35). In addition, the profiles of three participants with a rare autosomal dominant cerebellar disease were assessed (episodic ataxia type 2, EA-2). We found that participants with dyslexia were overall slower learners than controls in eyeblink conditioning. Further, they were the only group that had a reduced number of CRs in mediotemporal-dependent trace paradigm compared to the more cerebellum-dependent delay paradigm. Second, ADHD was found to be related to larger CR amplitude. Third, those with a comorbid condition learned faster and manifested CRs that were not well timed. Fourth, the cerebellar patients showed nearly no conditioning at all. Correlations between EBC and various neuropsychological domains (phonological processing, reading, spelling, arithmetic, executive functions, attention, and fine motor control) over all participants resulted in significant relations only for the delay paradigm: Increased amount of reading errors related with later peak latency and increased amount of self-corrections in fine motor control related with larger response magnitude. Within those who conditioned, relations emerged only for the trace paradigm: better spelling was related to larger response magnitude. These results do not lend support to the cerebellar hypothesis of dyslexia. On the contrary, dyslexia in its pure form seems to be related to a relative dysfunction of a larger hippocampal-cerebellar network. Further, larger responses in the ADHD group are suggested to result from their lowered responding threshold.