Sleep laboratory studies of patients complaining of insomnia have demonstrated discrepancies between subjective reports and electroencephalograph (EEG)-recorded measures. In our research studies on sleeping aids, 60% of the self-described poor sleepers who reported usual sleep latencies of at least 45 min did not meet the laboratory qualification criterion of a 30-min or longer sleep latency. To learn to predict who would qualify for our studies, we compared 30 laboratory-qualified poor sleepers (QPSs) with 30 laboratory-disqualified poor sleepers (DPSs) on subjective report, mood, and all-night sleep laboratory variables. QPSs had significantly lower sleep efficiency and total sleep time in the laboratory, but these differences were due to the longer sleep latencies (50.7 ± 27.8 min vs. 15.2 ± 6.1 min) of the QPS group. QPSs and DPSs differed significantly in their morning estimates of their laboratory sleep latencies; as a group, QPSs gave an accurate estimate (51.6 ± 27.8 min), but DPSs were significantly more likely to exaggerate their sleep latencies. Although we did not identify ways of predicting which poor sleepers would show sleep-onset insomnia in the sleep laboratory, we did find that, in this young, healthy population, there are poor sleepers who give an accurate report of a rather severe sleep-onset insomnia.Sleep-onset insomnia -difficulty in falling asleep -is a common sleep problem. In sleep laboratory studies, polysomnographic evaluations of patients complaining of insomnia have demonstrated discrepancies between subjective reports and electroencephalograph (EEG)-recorded sleep measures. Self-described insomniacs have been found to overestimate sleep latencies and underestimate total sleep time (