Sexual dysfunction often features as an outcome variable in community health surveys and epidemiological surveys. Key design imperatives for measures included in large scale, population-based surveys are acceptability, brevity and relevance to diverse sexual lifestyles. None of the available measures of sexual dysfunction are entirely suited to this task. We developed a new measure of sexual function for the third British National Survey of Sexual Attitudes and Lifestyles (Natsal 3). Items for the measure were derived from qualitative work from patients and community members. The draft measure was developed and validated using a general population sample (internet panel survey (n = 1,262)) and a clinical sample (patients attending sexual problems clinics (n = 100). Confirmatory factor analysis established that a 'general-specific model' had the best fit and was equivalent between general population and clinical samples (Comparative Fit Index = 0.963 Tucker Lewis Index = 0.951; Root Mean Square Error of Approximation = 0.064). The 17-item Natsal-SF is positively associated with the Female Sexual Function Index-6 (B = 0.572) and Brief Sexual Function Questionnaire for men (B = 0.705); it can discriminate between clinical and general population groups (OR = 2.667); and it has good test-retest reliability (r = 0.72). The Natsal-SF provides an estimate of the level of sexual function in the last year. By including items on distress about sex and sexual relationships, and by being relevant to all regardless of sexual lifestyle, it addresses some of the gaps in current measurement design.