Outcomes in speech perception following cochlear implantation in adults vary widely. Many studies have been carried out to identify and quantify factors that influence outcomes. This study adds a new dimension to pre-existing literature. Single-centre retrospective cohort study. University Medical Center Utrecht, the Netherlands. A total of 428 adults with bilateral severe-to-profound sensorineural hearing loss, unilaterally implanted between February 1988 and March 2014. Univariable and multivariable linear regression analyses were carried out to identify factors that may influence outcome after cochlear implantation. Consonant-vowel-consonant word scores were recorded pre- and post-implant and were used as outcome measure in two groups of patients (prelingually and postlingually deafened adults). As an added dimension, multiple imputation was implemented and evaluated to tackle 4% (17/407) missing data. For postlinguals, pre-implant speech perception score and age at onset of deafness are positive predictors and meningitis and otosclerosis as cause of deafness are negative predictors of post-implant speech perception. This model accounted for 26% of variance. For prelinguals, pre-implant speech perception score is the only strong positive predictor (β 0.524; P < 0.001). This model accounted for 31% of variance. Age at implantation was not a significant predictor in either group. Speech perception is predicted by pre-implant speech perception, age at onset of deafness and aetiology (meningitis and otosclerosis) for postlinguals and solely pre-implant speech perception for prelinguals. Age at implantation is of lesser importance in predicting speech perception outcome post-implant. Multiple imputation is a useful statistical technique when analysing incomplete data sets.