Purpose: The purposes of the study were to (a) investigate both explicit and implicit motor imagery ability (MIA) impairment after stroke, (b) examine predictive effects of clinical characteristics for MIA after stroke. Materials and Methods: Forty one patients with stroke (PwS) (mean age 59.41 +/- 10.19 years; %41 female) and 36 healthy participants (mean age 62.47 +/- 9.29 years; %47 female) completed Chaotic Motor Imagery Assessment-Hand Rotation for implicit MIA and Movement Imagery Questionnaire-3 (MIQ-3) and Box and Block Test (BBT) for explicit MIA. The severity of motor and sensory impairments were determined by the Fugl-Meyer Assessment-Upper Extremity (FMAUE) scores. The Turkish version of Motor Activity Log-28 was used to assess amount of use (AUS) and quality of movement in daily life. Results: Our results indicated that both implicit and explicit MIA (except kinaesthetic imagery of MIQ-3) in PwS were statistically impaired compared to controls (p < 0.05). The sensorimotor impairment level, amount of use and movement quality of the affected upper limb were found to be correlated with MIA in various degrees. Total motor scores in FMAUE and AUS were significant predictors of explicit MIA (p < 0.01). Additionally, explicit MIA scores of stroke subgroups were statistically different between severely and mildly impaired patients, in favour of mildly impaired group (p < 0.05). Conclusion: In conclusion, both motor impairment level and amount of daily use of upper extremity were found to be predictive factors for explicit MIA. Further investigation with brain imaging techniques is needed to explore the validity of these findings in establishing MIA.