This study investigates how cognitive and contextual characteristics of multiple-choice reading tasks impact empirical difficulty in a reading test. Results provide evidence of construct validity and are also relevant for teaching practice aiming at the teaching of reading strategies.