Background <p>Memory impairment is a common late adverse effect in survivors of many cancer types, but there are few studies of long-term survivors of testicular cancer (LTCSs). The aims of this study were to examine memory impairment in LTCSs both longitudinally and cross-sectionally, and to compare the cross-sectional findings with normative data.</p> Methods <p>A national sample of 753 LTCSs filled in questionnaires at means of 12 and 28 years after diagnosis. Memory impairment was reported by the Meta-Memory Questionnaire at 28 years. Longitudinal and cross-sectional associations between independent variables and memory impairment were examined with linear regression analyses.</p> Results <p>At 28 years mean memory impairment scores in LTCSs hardly differed from normative values in 10-year age groups from 40 to 79 years. Many variables were significantly associated with memory impairment in univariate regression analyses both cross-sectionally and longitudinally. In multivariable analyses total fatigue was significantly associated with memory impairment longitudinally, and with total fatigue and increased neuropathy cross-sectionally.</p> Conclusions <p>LTCSs at 28 years since diagnosis do not have increased level of memory impairment compared to normative data. In case of memory impairment complaints in LTCSs, several risk factors should be examined and some of them are amenable to interventions.</p>