Contrary to expectations, tonsillar carriage of S. pyogenes was similar among cases and genetically disposed and undisposed controls (0%, 1Á8% and 3Á3%, respectively; Table 1). Likewise, no significant differences in tonsillar or perineal carriage of S. agalactiae or S. dysgalactiae were detected, although carriage of these strains was more prevalent among cases. In contrast to analyses of data from microbiological cultures, self-reported tonsillectomy and frequent tonsillitis were more frequently reported among adolescents with psoriasis (11%) than among undisposed controls (2Á4%) (Table 1).Our results suggest that adolescents with psoriasis are not more likely to be carriers of S. pyogenes than their peers. Nonetheless, they are more likely to report recurrent tonsillitis, which is in line with our previous findings, but may be due to information bias. 5 Interestingly, the proportion of genetically disposed controls with recurrent tonsillitis was intermediate to that of cases and undisposed controls. The HLA-Cw*0602 allele may increase the risk of both psoriasis and streptococcal tonsillitis, which may explain this observation. 8 Lack of streptococcal carriers of S. pyogenes among cases of psoriasis runs contrary to our own expectations, and certain explanations may be plausible. The most obvious of these is that casesas our data suggestare more likely to undergo a tonsillectomy, thereby reducing the risk of future streptococcal colonization. In this context, a tonsillectomy may be a result of recurrent tonsillitis and may serve as a disease severity modifier. Lastly, S. pyogenes is known to create biofilms, leading to negative cultures even in the presence of colonization. Carefully taken samples and microbiome analyses may provide a clearer picture in future studies. Differences between our study and that of Mallbris et al. 8 deserve attention, but are likely explained by differences in time from disease onset.