Background: Pressure ulcers are common but preventable wounds. The United Kingdom National Institute for Health and Care Excellence Pressure ulcers: prevention and management clinical guideline suggests considering the use of risk assessment tools to support clinical judgement when assessing an individual's pressure ulcer risk. People deemed "at risk" may then receive frequent skin assessment to inspect for deterioration in skin condition and particularly for non-blanchable erythema. Both risk assessment and skin assessment inform use of subsequent preventative interventions including support surfaces (e.g., mattresses and overlays). There are many types of support surface from which to select and this is a challenging area of decision-making. My thesis identifies, syntheses and critiques available evidence on the prognostic value of risk and skin assessment on pressure ulcer prevention and the relative effectiveness of support surfaces in reducing pressure ulcer incidence. My thesis then addresses the overarching question: In pressure ulcer prevention what is the current evidence regarding (a) the prognostic performance of empirically-derived prognostic models in predicting pressure ulcer risk; (b) skin status descriptors that could be prognostic for pressure ulcer development; and (c) the clinical effects of support surfaces used for pressure ulcer prevention? Methodology and methods: An existing conceptual framework for pressure ulcer development supports the logic model proposed here which details how risk assessment, skin assessment and support surface use are linked in an overall model of pressure ulcer prediction and prevention. This thesis comprises four separate studies that explore existing evidence in each area outlined in the model: A prognostic model review (Chapter 4) identifying and evaluating risk assessment tools derived from multivariable analysis for pressure ulcer prediction (i.e., empirically-derived tools); A prognostic factor review (Chapter 5) and individual patient data meta-analysis (Chapter 6) investigating the prognostic value of specific skin statuses - particularly non-blanchable erythema - on pressure ulcer incidence; A prognostic factor review with individual patient data meta-analysis (Chapter 6) investigating the prognostic value of non-blanchable erythema on pressure ulcer incidence; An intervention review and network meta-analysis (Chapter 7) evaluating the relative effects of support surfaces in reducing pressure ulcer incidence. Within four separate reviews I conducted rigorous searches, screening, risk of bias assessment, data collection and meta-analysis appropriate to each review. I used GRADE, where available, to assess the certainty of evidence. Results: Findings from the prognostic model review of 24 included studies suggest that all 22 available, empirically-derived tools were developed using low-quality methods (i.e., risk of bias was judge as high or unclear in all cases). The prognostic factor review included forty-one studies (162,299 participants) investigating skin status-pressure ulcer incidence associations. Aggregate data suggest the odds of ulcer development are increased by 2.58 in those with non-blanchable erythema compared with those without (low-certainty evidence). Further analysis of individual participant data from four studies (3,223 participants) found moderate-certainty evidence that people with non-blanchable erythema have higher odds of developing a Stage 2 or above pressure ulcers than those without non-blanchable erythema (odds ratio 2.73, 95% confidence interval 2.10 to 3.52). The prognostic value of 14 other poor skin statuses with aggregate data meta-analyses is uncertain. The intervention review focused on support surfaces to prevent pressure ulcer incidence included 65 RCTs evaluating 14 support surfaces that can be provided for people at risk, or with non-blanchable erythema. Network meta-analysis found that powered active air and powered hybrid air surfaces are probably more effective than standard surfaces in reducing pressure ulcer incidence (risk ratio 0.42, 95% confidence interval 0.29 to 0.63; 0.22, 0.07 to 0.66, respectively). This is moderate certainty evidence. Conclusion: It is unclear if empirically-derived tools for pressure ulcer prediction can help distinguish between people at risk and not at risk due to the low-quality development process of these instruments and their performance requires further evaluation. People with non-blanchable erythema are more likely to develop a pressure ulcer than those without non-blanchable erythema. There is moderate-certainty evidence that powered active and hybrid air surfaces are more effective than standard options for preventing pressure ulceration. The comparative effectiveness of most support surface options for pressure ulcer prevention is still uncertain.