Working memory (WM) has been hypothesised to be impaired in attention-deficit/hyperactivity disorder (ADHD). However, there are few studies reported on tests measuring visuo-spatial WM (VSWM) in ADHD. Some of these studies used paradigms including episodic memory, others only used low memory loads. In the present study we used a VSWM test that has not been used previously in ADHD research. The sensitivity of the VSWM test and a choice reaction time (CRT) test was evaluated in a pilot study by comparing them to two commonly used tests in ADHD-research; the Continuous Performance Test (CPT) and a Go/no-go test, in children with and without ADHD. The groups differed significantly in performance on the VSWM test (P < .01) and CRT (P < .05) but not on the CPT (P > .1) or on the Go/no-go test (P > .1). The results from the VSWM and CRT tests were replicated in a larger sample of participants (80 boys; 27 boys with ADHD and 53 controls, mean age 11.4 years). The difference between the groups was significant for both the VSWM test (P < .01) and the CRT test (P < .01). The effect size (ES) of the VSWM test was 1.34. There was a significant age-by-group interaction on the VSWM test, with larger group differences for the older children (P < .01). Our results show that the VSWM test is a sensitive measure of cognitive deficits in ADHD and it supports the hypothesis that deficits in VSWM is a major component of ADHD.