Measuring the implementation fidelity (IF) or integrity of interventions is crucial, otherwise a positive or negative outcome cannot be interpreted. Direct and indirect methods of IF measurement tend to over-emphasize teacher behaviour. This paper focuses on IF measured by student behaviour collected through computers. Attainment was measured by the STAR test of maths (a computerized item-banked adaptive norm-referenced test). Implementation quality (IF) was measured by Accelerated Maths (AM) (an instruction-free personalized practice and progress-monitoring system in mastery of mathematics skills). Attainment data was gathered in the UK on 20,103 students in 148 schools, and of these implementation data on n = 6,285. Only a small percentage of pupils scored on five AM implementation indices at or above the levels recommended. Correlations between attainment and implementation indices were modest, but high implementation was positively correlated with high attainment. Socio-economic status did not appear to affect implementation or attainment. Implementation quality of AM is clearly a problem in the UK, and needs improvement. However, overall students still scored above average on attainment.