A pilot study is reported of a generic method for tutoring in mathematics, intended to be suitable for use by peers, parents and volunteers in a wide range of applications. Thirty children aged 9-10 years of below average mathematical ability were randomly allocated to experimental or control conditions. Experimental tutees (n=17) were tutored in mathematical problem solving at home by their parent(s) using the method, while control children (n=13) received traditional maths problem homework. Pre- and post-test assessment of both groups involved a criterion-referenced mathematics test in parallel forms and a scale of attitudes to mathematics. Experimental tutors completed a pre-test questionnaire on attitudes to mathematics and home-school links, and experimental tutees and tutors engaged in a post-test debriefing interview. On the attainment test, the experimental group gained significantly, while the control group did not. Male tutees gained more than females. No significant pre-post differences were evident on the tutee attitude questionnaire. However, interview feedback from both tutees and tutors was generally positive. Given the brevity of the pre- to post-test interval, the finding of positive differences in attainment was considered encouraging. Recommendations for future research were made.