The objective of the present study was to assess alexithymia by means of the Toronto Alexithymia Scale (TAS-20) and The Emotion Protocol (EP) in a group of refugees. Eighty-six subjects were willing to participate. At last follow-up, 33 non-PTSD and 22 PTSD subjects had complete data. Subjects with PTSD had higher scores on the TAS-20 (F = 4.314, df = 77, p = 0.041), but on the subscale level, this was significant only with regard to Factor I, difficulties identifying feelings (F = 5.316, df = 77, p = 0.024). TAS Factor I and to a lower extent TAS Factor II (difficulties naming feelings) were significantly associated with the self-rated presence of dysphoric affects. At follow-up, an increase in TAS Factor I score was associated with increased prevalence of self-rated symptoms of PTSD, but not depression. Decrease in prolactin was associated with significant increase of TAS Factor I (rho = -0.396, n = 54, p = 0.003). The present study indicates that alexithymia as measured by TAS-20 is indeed associated with symptoms of PTSD. This association is almost exclusively explained by the TAS Factor I subscale and is in turn associated with a high level of self-reported dysphoric affect. The longitudinal inverse correlation with prolactin points to the possibility of an underlying disturbance in serotonergic and/or dopaminergic systems. The results thus indicate that secondary, or post-traumatic, alexithymia is a measure of suppressed or warded-off negative affects.