who reported that risk of prostate cancer follow-ing diagnosis of nonmelanoma skin cancers was reduced to agreater extent in those living in sunny countries (specificallyAustralia, Singapore and Spain) than in those living in lesssunny countries. The report of Tuohimaa et al.isproblematic.Australia, represented in this analysis only by the New SouthWales Central Cancer Registry (NSWCCR), cannot have beenincluded in the results Grant and Garland cite. The NSWCCRdoes not and has never registered basal and squamous cell car-cinomas of the skin. Moreover, the results of Tuohimaa et al.cannot be used to make valid inferences about associationsbetween sun exposure and cancer. The reasons are well sum-marized in an IARC Working Group report:‘‘The incidence of second cancers in individuals is elevatedby several known and unknown mechanisms, includingcommon aetiological factors and predispositions, andinfluenced by possible biases in the ascertainment of sec-ond cancers in the cancer registries. The net direction ofthese influences will mostly be in the direction of elevatedoccurrence of second cancers, against which a possibleeffect of sunlight and vitamin D (manifest through theoccurrence of NMSC [non-melanoma skin cancer]) couldbe difficult to detect with the type of cohort design thathas been used in these studies and the lack of informationon known risk factors in many of the studies.’’