The literature on teaching/learning of negotiation skills presents two assumptions. First, planning for negotiations – or improving one’s planning – best predicts more effective negotiation practice. This suggests the need to tie planning to assessment, particularly as developing negotiation plans is the most ‘authentic’ assessment task a student can perform. Second, ‘inauthentic’ assessment tasks for knowledge acquisition – like multiple choice exams – are unhelpful for learning negotiation skills. We address the relationship between ‘authentic’ and 'inauthentic' assessment in student learning of planning. The authentic tasks are three negotiation plans students prepared for separate simulations. We analyzed the marks (scores) 639 students from four successive semesters received for their plans before (Plan 1) and after (Plan 2) they sat a common multiple-choice exam. We initially compared students’ marks on the three different planning tasks (testing differences of means) to verify than any apparent relationships broadly complied with theoretical expectations. We found that preparation for a knowledge acquisition test of core concepts lifts planning performance between Plans 1 and Plan 2. Because the level of challenge increases with each successive plan – in terms of length, complexity and sophistication – marks did not increase between Plans 2 and 3. This suggests that the increased level of challenge offset any improvement in ‘doing plans’. We then subjected any meaningful pattern of differences between planning results to factor analysis to explore the underlying structure of any identified patterns. Our research found major underlying differences between plans before and after acquisition of core concepts.