The Mental Adjustment to Cancer (MAC) scale was validated on a heterogeneous French sample of 317 cancer patients. Internal consistency was satisfactory for the original subscales (α coefficients=0.62–0.80), except for the Fatalism subscale (α=0.40). The intercorrelations of the subscales and the correlations between the subscales and Anxiety and Depression criteria were congruent with the values reported in the literature. Multidimensional Scaling revealed three positive and three negative subsets of items revealing adjustment to cancer. Congeneric factor analysis of the subsets was performed with LISREL 8.3 and only three of them (after discarding certain items) were retained: Fighting Spirit (FS) Hopelessness/Helplessness (HH) and Anxious Preoccupation (AP). A confirmatory hierarchical factor analysis on the 21 items included showed that FS measured positive adjustment to cancer and HH and AP measured negative adjustment. A differential adjustment hypothesis was proposed in order to explain the stability and instability of the measures of the diverse constructs. The three revised subscales showed the same validity pattern as the corresponding original scales, but the magnitude of the correlations was considerably improved with respect to the original subscales. The practical and the theoretical importance of FS, HH and AP are emphasized. Copyright © 2003 John Wiley & Sons, Ltd.