Students and researchers with no or limited experience in setting up BOLD fMRI studies in terms of paradigm (task) design After this presentation, the audience will know (i) what BOLD fMRI can and cannot tell you of brain activity, (ii) how this is affected by the BOLD hemodynamic response function, (iii) what paradigm design comprises in BOLD fMRI studies and why it is so important to correctly design your experiment, and (iv) the different types of experimental designs commonly used.FMRI experiments generate 4-dimensional datasets, i.e. series of volumes acquired in time. From these, we aim to extract specific patterns in time and space, to ultimately reveal neuronal activity information about the brain. To do this, first a pattern has to be brought into the system, which is typically accomplished by presenting a subject with a task (‘paradigm’). The aim of paradigm design is to choose and optimize a set of stimulus types and their organization in time. The final design completely determines the interpretation of the end result: statistical activity (t-) maps. The choice of a paradigm depends on the objective of a study and your hypothesis. A very specific hypothesis (what brain functions? which regions?, importance of timing?) greatly benefits the interpretation of the end result. Important to note is that the fMRI brain response is an relative measure and thus has no meaning independent of the control condition. Therefore, to evaluate the brain’s response to an experimental manipulation we will always, at least, contrast one experimental condition to a control condition. Conditions can be organized according to various schemes, such as a subtraction, conjunction, factorial or parametric design, and their timing (block, or event-related design). The choice of a particular scheme and control condition is one of the most difficult issues and will involve optimizing the scheme in terms of validity (am I measuring what I thing I am measuring?), sensitivity (can I measure what I want to measure?), and reliability (will my results be reproducible?). Lastly and most importantly, your subjects should be able to perform the tasks with consistent task performance (or find ways to measure/control for subject performance), especially when comparting healthy subjects and patients. A carefully designed experimental fMRI paradigm that maximises the validity, reliability, sensitivity, and proper interpretation of the brain imaging results.The main goal of paradigm design is to isolate the activity related to mental processes of interest in a BOLD fMRI experiment. This depends above all on your hypothesis and question. A wide variety of experimental design types can be employed with identifiable pro and cons (block vs event-related, factorial, parametric etc.). An optimal scheme can be estimated based both on the temporal features of the invoked processes, and on the mathematical characteristics of the scheme in terms of efficiency, time between stimuli, collinearity between conditions, and degrees of freedom in the statistical analysis. As such, the statistical image analysis plays a significant role in paradigm design. Many of the important design issues can be dealt with before any fMRI scans are acquired, based on prior knowledge and the ability to isolate and identify the processes of interest (but also those that are hidden), technical constraints, mathematical properties of the stimulus scheme, and subject performance (patients vs controls).