<p>The growth of microfluidics has greatly increased the demand for microfluidic products. CO₂ lasers are being extensively employed to manufacture microfluidic devices using PMMA because of the flexibility, versatility, and cost-effectiveness involved. However, defining the relationships among the input parameters and responses, and getting the optimum results are challenging. Techniques of design of experiments like the Taguchi method, response surface methodology (RSM), and 2-level factorial design help to achieve optimum results. In this work, RSM based on the central composite design was employed to investigate the effect of the laser power (1.5, 3.0, and 4.5&#xa0;W), scanning speed (10, 15, and 20&#xa0;mm/s), and pulse rate (800, 900 and 1000 pulses per inch) on the microchannel depth, width, aspect ratio, and heat-affected zone (HAZ), during CO₂ laser fabrication of microchannels on PMMA substrates coated with a 500&#xa0;nm layer of 99.95% pure aluminium. A total of 54 experiments were conducted, full quadratic models were developed, and multi-objective optimization was done. The optimization results were validated experimentally, and a repeatability test was performed using the optimum conditions. Using analysis of variance (ANOVA), the models were tested at a 95% confidence interval, and the significance of the parameters was determined. The laser power is the most significant factor followed by the scanning speed and pulse rate. The models are adequate with small absolute prediction errors of 2.21%, 1.54%, 2.53%, and 2.87% for the microchannel width, depth, aspect ratio, and HAZ respectively. For every response, the repeatability result is high with a small standard deviation of 1.809, 1.358, 0.022, and 1.949 of the microchannel width, depth, aspect ratio, and HAZ respectively. The models can be used by manufacturing engineers to efficiently predict the microchannel characteristics.</p>