OBJECTIVE:To examine the distribution, consistency, and correlates of sleep time perceptions in primary insomnia sufferers and normal sleepers across settings (laboratory and home). METHODS:Fifty-two middle-aged and older insomnia sufferers and 49 matched normal sleepers underwent multiple nights of laboratory and home polysomnography (PSG) and provided concurrent subjective estimates of their sleep. Descriptive statistics, cluster analyses, and inferential tests were used to examine the nature of sleep time perceptions, identify subgroups with distinctive patterns such as perceptions, and explore factors that may contribute to these subgroups' distinctive sleep perceptions. RESULTS:Consistent with previous studies, the insomnia sufferers, as a group, showed a greater tendency to underestimate the time they slept than did the normal sleepers, but the accuracy of sleep time perceptions varied widely within each sample. Correlational analyses showed that study participants in each sample showed a moderate level of consistency in how accurately they estimated sleep time across nights. Subsequent cluster analyses conducted to identify subgroups with distinctive patterns of sleep time perceptions yielded four subgroups within the insomnia sample and three subgroups within the normal sleeper sample. Although the sleep setting significantly affected sleep perceptions for one insomnia subgroup, both laboratory and home objective-subjective sleep comparisons provided similar impressions regarding the relative accuracy of each subgroup's nocturnal sleep/wake perceptions. Finally, varying patterns of subgroup differences were noted across several common psychometric measures. CONCLUSIONS:The underestimation of sleep time is not a generic characteristic that separates all insomnia sufferers from normal sleepers. Sleep setting, personality traits, and perhaps constitutional factors appear to influence perceptions of sleep and wake time duration.