The Wechsler Intelligence Scale for Children – Third edition (WISC-III) has been widely used to assess the intellectual abilities of children with hyperkinetic disorders by both researchers and practitioners. The characteristic features of the WISC-III profile of this clinical group have been described by several authors (Anastopoulos et al., 1994; Assessmany et al., 2001; Snow and Sapp, 2000; Mayes and Calhoun, 2006; Ek et al., 2007). However, the issue of the WISC-III construct validity in this particular clinical group was scarcely addressed. It remains possible that in hyperkinetic children with poor attention this long and attention-demanding test measures other constructs than in general population. The objective of this study was to examine the structure of intelligence in primary school children diagnosed with hyperkinetic disorders. 29 children (23 boys and 6 girls) aged seven to ten years, participated in the study. All the children were diagnosed with a hyperkinetic disorder, and ten of them had a comorbid diagnosis of specific learning disabilities. The participants of the study completed the Lithuanian version of the Wechsler Intelligence Scale for Children – Third edition (WISC-III) (2002). The Confirmatory factor analysis confirmed the construct validity of WISC-III when used in hyperkinetic children: one factor model (general intelligence) fits the data satisfactorily, while two factor (Verbal IQ, Performance IQ) and four factor (Verbal Comprehension, Perceptual Organisation, Freedom from Distractibility, Processing Speed) models fit the data very well. However, the exploratory factor analysis revealed a different factor structure without the Processing Speed factor, more similar to WISC-R. It also showed that Object Arrangement is a problematic subtest when used with hyperkinetic children as it does not fit with any of the three factors. We suggest that its performance could be determined by emotional variables rather than by cognitive ones. The WISC-III profile of hyperkinetic children as a group was characterized by the low Freedom from Distractibility score and ACID profile (low Arithmetic, Coding, Information and Digit span scores). The mean scores of these subtests and factors were significantly lower than the standardization sample mean, which is 10 for subtest scores and 100 for factor scores: M = 8.4, t = –3.215, p < 0.01 for Information; M = 8.5, t = –5.998, p < 0.01 for Coding; M = 8.4, t = –3.194, p < 0.01 for Arithmetic; M = 6.9, t = –5.998, p < 0.01 for Digit Span; M = 86.17, t = –5.393, p < 0.01 for Freedom from Distractibility. These peculiarities of hyperkinetic children’s WISC-III performance are explained by difficulties in attention, short-term memory, sequential auditory processing and factual knowledge. Satisfactory results of Similarities, Picture Completion and Block Design suggest relative strengths of using concepts, ability to analyse and synthesize information, and fluid intelligence. It has been proposed that comorbid specific learning disabilities may affect the cognitive performance of hyperkinetic children. For this reason, the effect of a comorbid learning disability was also assessed. The results showed that, although the intellectual abilities of hyperkinetic children with a comorbid learning disability are lower, the comorbid learning disability does not change the hyperkinetic children’s intelligence profile.