BACKGROUND: Human studies of dietary fat as a possible risk factor for cutaneous malignant melanoma (CMM) and non-melanoma skin cancer (NMSC)--principally basal cell carcinoma (BCC) and squamous cell carcinoma (SCC)--have produced inconsistent results. We had the opportunity to examine the association concurrently for all three types of skin cancer in a population-based study in Tasmania, Australia, involving 652 cases of CMM, BCC and SCC and a common set of 471 controls. METHODS: Histopathologically-confirmed cases of CMM, BCC and SCC were ascertained from the Tasmanian Cancer Registry (TCR), and controls were selected at random from the state's electoral roll. We compared subjects categorised by thirds of dietary fat intake score measured by the 'Dobson short fat questionnaire', with logistic regression models that adjusted for age, sex, skin type and usual sun exposure. We then followed all subjects for 56-80 months until 31 August, 2004 for a new NMSC using record linkage with both the TCR and the Births, Deaths, and Marriages registry. Incidence rates were calculated and ratios of rates were estimated using Poisson models. RESULTS: Relative to subjects in the lowest fat intake category, the odds ratios (OR) comparing cases and controls were OR = 0.76 (95% CI: 0.56-1.03) for medium fat intake, and OR = 0.62 (95% CI: 0.45-0.85) for high fat intake, with a significant (p < 0.01) trend of reduced odds ratio with higher category dietary fat intake. Among cases, the incidence rate ratio (IRR) relative to those with lowest fat score was IRR = 0.72 (95% CI: 0.50-1.03) for medium fat intake, and IRR = 0.82 (95% CI: 0.56-1.20) for highest fat intake (linear trend p = 0.30). CONCLUSION: Using the same dietary instrument with two study designs in the same Caucasian population, we found no evidence that high fat intake increases the risk of developing melanoma or non-melanoma skin cancers. Instead, our results suggest a risk reduction for high fat intake.