The current study served to replicate the finding that extensive derived relational responding fluency training using the SMART (strengthening mental abilities with relational training) method can enhance intelligence quotient scores, while controlling for baseline levels of intelligence and attentional skills. Two separate groups of children based on school enrolment, aged 10–11 years, were assigned to a control or treatment condition en masse (i.e., not individual random assignment). All participants were administered the Test of Everyday Attention at baseline and completed the Wechsler Abbreviated Scale of Intelligence (WASI-II) at baseline and follow-up. Supervised training in derived relational responding fluency was administered at regular intervals over several months for the intervention group only. Results showed that there was a significant increase in IQ scores from baseline to follow-up for the intervention group only. However, controlling for baseline intelligence scores or any of three separate sub-indices of attentional skills eliminated this effect. This outcome serves as an indicator of the importance of controlling for baseline factors in intervention studies and the random allocation of participants to treatment and control groups.