Luthe has contended that autogenic training "promotes processes of self-regulatory re-adjustment for a variety of disturbed autonomic functions" (Luthe, 1970, p. 125). To test the hypothesis that this re-adjustment would be expressed in a convergence of bilateral differences in autonomic variables during or after autogenic training, 12 subjects were exposed to a modified autogenic training protocol in a within Subjects design which contrasted three periods: pre-relaxation arousal, autogenic relaxation, and post-relaxation arousal. There was no convergence between left and right measures of skin conductance, pulse volume, or skin temperature either during or after autogenic training. There were no left-right differences in the variables measured bilaterally, either across subjects in any period or across periods. Conductance, heart rate, and self-reported arousal were reduced, temperature was increased, and pulse volume was unaffected by the autogenic training. The results supported Lacey's (1967) conception of the autonomic nervous system being a multidimensional set of loosely connected, complexly interacting systems.