25 project reports from the second semester, labelled "Theory, Model, and Experiment" of the Natural Sciences Basic Programme at Roskilde University were analysed. The analysis concentrated on the elements of experimental work performed by the students. The assessment tool used was formulated in broad categories of elements of practical work, common to all the natural sciences in order to embrace all different traditions of "good scientific performance". The majority of the reports revealed the students' engagement in most stages of experimental work, including the formulation of the contribution of the practical work to the solution of the problem under study. The Chemistry-dominated problems seemed not to be better off in this respect than projects from other areas of the natural sciences. Experimental training of general relevance to the study of Chemistry actually takes place for the majority of students in their second first-year project. However, since the assessment tool was formulated in general terms, it does not discriminate between practical work which is central in most chemical investigations, and that which is of less importance.