Background Cyberchondria is defined as the increase in health-related anxiety or anguish associated with excessive or repeated online searches for health-related information. Our objective was to cross-culturally adapt and validate the CSS-12 scale for Peruvian Spanish speakers, to determine whether the Bifactor model works as well in our population as in previous studies’ and to explore whether the Bifactor-ESEM is a more suitable model. Methods We performed a cultural adaptation using the Delphi method and a validation study on medical students between 2018 and 2019. Reliability was evaluated by using Cronbach’s alpha (α) and McDonald’s omega (Ω) for internal consistency, and Pearson’s r and intraclass correlation coefficient (ICC), for test-retest reliability. We evaluated construct validity by contrasting four measurement models for the CSS-12 and the convergent validity against health anxiety. Results The Spanish CSS-12 showed excellent reliability (α = .93; Ω = .93; ICC = .93; r = .96). The Bifactor ESEM model showed the best fit, supporting a unidimensional measure of the general cyberchondria. This measure was positively associated with health anxiety (r = .51). Conclusions The Spanish CSS-12 provides a valid and reliable unidimensional measure of cyberchondria, which is distinguishable from the more general health anxiety. This can be applied to similar populations and future research. The Bifactor-ESEM model appears to offer a more accurate and realistic representation of the multifaceted nature of cyberchondria. We provide a free-to-use form of the Spanish CSS-12 as supplemental material.