The purpose of this multi-sample study was to examine the psychometric characteristics, factor structure, and measurement invariance of the hard-copy and web-based versions of a measure of sport imagery ability, termed Sport Imagery Ability Measure (SIAM). In the first sample, Spanish athletes (N = 274, 161 men, 113 women, Mage = 21.91, SD = 6.67) completed a hard-copy version of the SIAM. A newly developed web-based version of the SIAM was cross validated in an independent group (N = 266, 147 men, 119 women, Mage = 25.93, SD = 9.84). A small group of participants (n = 16) completed both versions. Exploratory structural equation modelling and confirmatory factor analysis of the data from the hard-copy version showed that a 3-factor (i.e. generation, feeling, and single senses) solution of the SIAM reached satisfactory fit indices. The 3-factor solution showed good fit to the data obtained through the web-based version of the SIAM. Multi-group comparisons provided support for measurement invariance across gender and competitive level. Evidence of full invariance of factor loadings was obtained for both formats of administration indicating that the factor structure was held across groups, while partial invariance was seen for item intercepts indicating inequality in the intercepts of some indicators.