For the last three decades, male and female sexual responses have been conceptualized as similar, based on separated and sequential phases as proposed by the models of Masters and Johnson (1966) and Kaplan (1979) model. However, there is a growing debate around the need to conceptualize female sexual response and the classification of sexual dysfunction in women, in view of the upcoming editions of the DSM and ICD. The aim of this study was to test, using structural equation modeling, five conceptual, alternative models of female sexual function, using a sample of women with sexual difficulties and a sample of women without sexual problems. A total of 1993 Portuguese women participated in the study and completed a modified version of the Female Sexual Function Index. Findings suggested a four-factor solution as the model that best fit the data regarding women presenting sexual difficulties: (1) desire/arousal; (2) lubrication; (3) orgasm; (4) pain/vaginismus. In relation to sexually healthy women, the best model was a five-factor solution comprising of (1) desire; (2) arousal; (3) lubrication; (4) orgasm; and (5) pain/vaginismus. Discriminant validity between factors was supported, suggesting that these dimensions measure distinct phenomena. Model fit to the data significantly decreased in both samples, as models began to successively consider greater levels of overlap among phases of sexual function, towards a single-factor solution. By suggesting the overlap between pain and vaginismus, results partially support the new classification that is currently being discussed regarding DSM-5. Additionally, results on the relationship between sexual desire and arousal were inconclusive as sexually healthy women were better characterized by a five-factor model that considered the structural independence among these factors, whereas women with sexual difficulties better fit with a four-factor model merging sexual desire and subjective sexual arousal.