AbstractProcess schizophrenics, reactive schizophrenics, neurotics, and alcoholics (N= 80) estimated the duration of a dark dot that followed an auditory warning signal of variable duration. The duration of the warning signal constituted foreperiod (FP). The effect of FP duration on time estimation (TE) was an increasing one for allSs, and schizophrenics were indistinguishable from nonschizophrenics. Similarly, there was no group difference with respect to the effect of preceding foreperiod (PFP). Prior time information (TI) in terms of the correlation between the pitch of the warning signal and FP duration eliminated the effects of FP and FP change for schizophrenics. As for neurotics, such prior TI accentuated the effect of decrease of FP duration and attenuated the effect of increase of FP duration. The multi‐level conceptualization of time is proposed in order to establish the distinction between regression in time and regression of time. Neurosis is regarded as involving regression in time; schizophrenia is regarded as involving regression of tim