Background Measurement is essential for methods of instruction to be successful. Having an instrument that is reliable and validated in the given setting is vital. The primary goal of this research project is to validate the instructional practice scale (IPS) for university instructors in the Ethiopian context. Methods By implementing a cross-sectional descriptive survey research design, 1,254 participants across four public universities – Arbaminch, Dilla, Wachamo, and Jinka representing the first, second, third, and fourth generations, respectively were randomly selected and participated. The data was split in half and underwent an exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Results The three components of the EFA were alternately filled with seventeen items that satisfied certain standards, had a loading value of > .5, and a Cronbach alpha of ≥ .874. The factors identified in the EFA have been confirmed to be the thirteen items with loading, Cronbach alpha, Raykov’s rho coefficient (rho_A), composite reliability (CR) value of >.7, and Average Variance Explained (AVE) value of > .5. Tests of measurement and structural models showed a good fit. The Fornell-Larcker criterion, which is employed in discriminant validity analysis, demonstrates that the square root of the AVE for each construct is higher than the correlation it exhibits with other constructs. The correlations’ heterotrait–monotrait (HTMT) ratio is getting close to zero, and there isn’t one in the confidence interval at the.05 significance level. Both guaranteed strong discriminant validity. Conclusion For university instructors, the 13 items generally have powerful psychometric properties. The three subsections of the instruction practice scale—planning (4 indicators), delivering (4 indicators), and assessment (5 indicators) — are meant to measure the instructional practice effectively. Implications of the findings were further discussed.