Considering the reported lack of interest in the STEM-domain and the consequential difficulties in recruiting talented and interested young academics, the development of effective enrichment measures is indispensable. This requires a precise picture of students’ interests. The paper presents an approach to characterize interest profiles in explicitly science-related activities. Adapting Holland's RIASEC-model, an instrument was developed and tested which allows the description of interest in activities along Holland's dimensions (and a seventh dimension networking) within the confined science domain. The findings of a study with N = 247 students (age cohorts 12–19 years) uncovered interest differences for the environments school, enrichment, and (prospective) vocation. The mutual importance of the performed activity and the environment the activity is performed in is confirmed by a cross-classified model. Contrasting different subgroups revealed multiple results, e.g., girls showed more interest in artistic and social activities within the science domain. High achieving students showed more interest in science-related activities in all dimensions. In conclusion, using our adapted model, students’ interest structure can be described in a differentiated manner. This could lay the foundation for further analyses of students’ interest profiles and thereby contribute to future development of effective and congruent enrichment measures, thus enhancing interest in science.