Background: A wide range of studies has offered suggestions why genetics is difficult and some of their key findings are summarised. Underpinning all of this is the way the brain works when handling information. The limitations of working memory capacity offer an interpretation of these difficulties. Purpose: The aim is to confirm that working memory capacity (and the related concept of field dependency) controls performance in understanding genetics and whether it is possible to improve performance by changing the teaching approach to mininise overload. Programme description: The curriculum in Taiwan in genetics is outlined briefly. A wide range of measurements were made. Using a diagnostic test of understanding of underpinning ideas, the key areas of weakness were detected before the pupils started the course. Sample: Stage 1: 141 students in Taiwan, aged 13, boys and girls, drawn from a cross-section of Taiwanese pupils at this age, following their first course in genetics. Stage 2: 361 students, drawn from a cross-section of Taiwanese pupils at the same age, and divided into two groups (experimental-control) which both encompass the same ability range. Design and methods: Stage 1: test of pre-knowledge using structural communication grids, applied before the course commenced; working memory capacity using the figural intersection test; extent of field dependency using the group embedded figure test; understanding of genetics was measured at end of course; school test data collated. Stage 2: following a completely revised approach, performance in genetics was measured using traditional school tests and a word association test. Student perspectives were measured. Results: The test of pre-knowledge revealed key areas of difficulty. In addition, it was found that working memory capacity and extent of field dependency both correlated extremely highly with all measures of performance. Given that it has been established that working memory capacity controls performance, working memory demand explains why genetics is difficult. It was found that re-structuring the teaching approach to minimise mental overload brought about a very marked improvement in performance. Conclusions: The findings suggest that it is possible to reduce difficulties in understanding genetics by means of teaching re-design to minimise potential working memory overload, with concomitant improvements in learner confidence. (Contains 9 tables and 3 figures.)