How well the learning transfers from one situation to another is one of the most critical issues in education. Current educational theories emphasize the positive effects of providing learners with control over their own learning (Williams, 1996). Learner control increases learners' willingness to invest effort, which may yield a higher transfer performance. However, research often shows that learners may opt to omit important instructional material, hampering learning. Hence, instruction should integrate some degree of learner control, rather than including full learner or system control. Moreover, task variability is believed to enable learners to distinguish task-relevant information from irrelevant context, facilitating schema (knowledge structures) construction and the adaptation to new instances, enhancing transfer (Gick & Holyoak, 1983). Therefore, good instruction should ensure task variability and provide an appropriate amount of freedom to learners. In the genetics domain, the effects of shared control (a combination of system and learner control) and of presenting learners with tasks varying in surface features on transfer performance are studied.Eighty-six students participated in this study. A randomized 2x2 factorial design with variability (low, high) and control (system, shared) was used. A task-database contained 162 inheritance tasks with species (human, animal, plant) and trait (colour, shape, length) as surface -irrelevant- features, and crossing-type (combining homozygous and heterozygous parents) as structural -relevant- feature. Learners worked on 12 training tasks. Species and trait were invariable in the low-variability conditions and differed from the preceding task in the high-variability conditions. The system-control condition presented one task. The shared-control condition preselected three tasks, from which participants made the final selection. The transfer test contained 12 (6 near and 6 far transfer) tasks. Transfer test reliability was ? = .83. Mental effort was measured after each training (? =.97) and transfer problem (? =.95 ) with a one-item 7-point rating-scale.Results show interactions on both transfer (F(1,82)= .310, p< .05) and task involvement (F(1,82)=5.085, p< .05), indicating that shared-control yields a better transfer performance and task involvement for high-variability as compared to system-control, which yields better results in combination with low-variability. Acording to Paas et al.'s (2005), when task involvement is high, more mental effort is likely to be invested in learning, which is likely to result in a higher transfer performance. Transfer scores and mental effort scores are first standardized and the z-scores are entered into the formula: Task Involvement = Z Transfer + Z Mental Effort / ?2 Results of this study indicate that students who are provided with shared control, profit more from this control if the surface features of the learning tasks to choose from are different than those in the previous task. No effects on mental effort were found. The same amount of mental effort yielded higher transfer in the high-variability with shared-control condition, which may indicate that a higher proportion of the mental effort was invested for genuine learning. These findings provide instructional implications. Learner control should be provided when task surface features are different from each prior task. However, the optimal amount of control, especially when structural features highly vary, requires further investigation, since they may both overwhelm learners when provided together.Gick,M., & Holyoak,J. (1987). The cognitive basis for knowledge transfer. In S. Cormier and J.Hagman (Eds.), Transfer of learning: Contemporary Research and Applications (pp.9-46). San Diego, CA: Academic Press. Paas,F., Tuovinen, J., van Merriënboer,J.J.G., & Darabi,A. (2005). A motivational perspective on the relation between mental-effort and performance: Optimizing learner involvement in instruction. Educational Technology, Research and Development, 53(3),25-34. Salomon,G. (1983). The differential investment of mental effort in learning from different sources, Educational Psychologist, 18(1),42-50. SSCI journal