Students' images of science and scientists are generally assumed to influence their related subject choices and aspirations for tertiary education within science and technology. Several research studies have shown that many young people hold rather stereotypical images of scientists, making it hard for them to see themselves as future scientists. Adolescents' educational choices are important aspects of their identity work, and recent theories link individual choice to the perceived match between self and prototypical persons associated with that choice. In the present study, we have investigated images of scientists among the segment of the upper secondary school students (20% of the cohort) from which future Danish scientists are recruited. Their images were rather realistic, only including vague and predominantly positive stereotypical ideas. With a particular Science-and-Me (SAM) interview methodology, we inquired into the match between self- and prototypical-scientists (N?=?30). We found high perceived similarity within a core of epistemological characteristics, while dissimilarities typically related to a social domain. However, combining interview data with survey data, we found no significant statistical relation between prototype match and aspirations for tertiary education within science and technology. Importantly, the SAM dialogue revealed how students negotiate perceived differences, and we identified four negotiation patterns that all tend to reduce the impact of mismatches on educational aspirations. Our study raises questions about methodological issues concerning the traditional use of self-to-prototype matching as an explanatory model of educational choice.