The idea that blood groups may be associated with particular kinds of disease is not new, but interest in it has recently been rearoused by statistically significant reports of such associations. The most striking finding is the relatively high frequency of blood Group O in patients with duodenal ulcer, but a number of others have been recorded (Lancet, 1958). We felt it might be useful to see whether any such association could be found in skin diseases, and we selected psoriasis because of its clear-cut entity and its hereditary tendency. Previous investigations have shown marked disagreement. Klövekorn and Simon (1927) found no correlation between blood groups and psoriasis and also other skin diseases. Nardelli (1928) on the basis of 42 patients came to the same conclusion, though his figures do suggest some preponderance of Group O compared with the normal population. Poehlmann (1928) also found