Mathematics is a key subject which students cannot avoid if they have to lead a bright future. Despite the introduction and implementation of different teaching methods and strategies suggested by researchers the achievement of students in mathematics at school level has persistently been poor, hence the need to explore the influence of different instructional approaches. The purpose of this study was to find out the extent to which instructional practices influenced students' achievement in secondary school mathematics. This entails the use of Behavioral Objective-Based (BOB), Peer Instruction enriched with Concept Tests (PICT), both BOB and PICT (BO-PICT) and Conventional Instruction (TI). The theoretical framework which guided this study was social constructivism theory. The study adopted quasi-experimental, utilizing pretest posttest non-equivalent group design. The target population for the study was all the 3056 form three students in the 38 secondary schools in Mumias sub-county. Disproportionate stratified sampling and simple random sampling was used to select 327 students who took part in the study. Mathematics Achievement Test (MAT) was used to collect data. Data was analyzed using both descriptive and inferential statistics. Findings revealed that there was a significant difference in achievement, between all Experimental groups (BOB, PICT and BO-PICT) and the control group in favor of the Experimental groups. There was a significant difference between means of groups that used BOB and BO-PICT as well as PICT and BO-PICT. Those ones who used BOB and PICT were found to have similar achievement in the learning of mathematics. Based on the findings, BOB, PICT and BO-PICT were found to be viable instructional strategies that could promote performance in mathematics.