Sierra Leone’s new Senior Secondary School curriculum seeks to move mathematics teaching beyond examination-driven content coverage towards problem solving, reasoning, application, communication and interdisciplinary learning. This study protocol describes a quantitative cross-sectional explanatory study designed to examine self-reported implementation fidelity of the new Mathematics for STEAMM curriculum among mathematics teachers in Eastern Sierra Leone and to identify teacher-reported conditions associated with variation in implementation. The planned study covers 20 senior secondary schools distributed across Kenema, Kono and Kailahun Districts, with a working allocation of 40 mathematics teachers and 200 senior secondary students. The teacher instrument measures self-reported implementation fidelity, teacher preparedness, professional development, resource adequacy, classroom conditions, instructional workload constraints, assessment alignment and school leadership support. Student responses provide supplementary evidence on classroom implementation as experienced by learners. Analysis will begin with descriptive and reliability assessment, followed by bivariate analysis, an unconditional random-intercept model to estimate between-school variance and the intraclass correlation coefficient, and a parsimonious multivariable random-intercept model. Because the design contains only 20 schools and a small number of teachers per school, inference will emphasise effect sizes, confidence intervals, variance-component uncertainty and a cluster-adjusted sensitivity analysis. The protocol is anchored in Rogan and Grayson’s Theory of Curriculum Implementation and is designed to provide contextually relevant evidence on the teacher and organisational conditions associated with curriculum enactment in Eastern Sierra Leone.