The new combination of amlodipine (AML) and celecoxib (CXB) has been recently approved for co-treatment of hypertension and osteoarthritis in the form of fixed-dose combined tablets. For the first time, an innovative HPTLC method was developed using quality by design approach for determination of amlodipine and celecoxib in presence of process related-impurity; 4-methyl acetophenone in their combined dosage form. Also, the molecular modeling techniques were applied to investigate the inhibitory effect of 4-methyl acetophenone on the human carbonyl reductase enzyme (type I) and the potential health hazard which makes essential to control its level in the dosage form. The chromatographic conditions were optimized by applying quality by design principles during the development phase using experimental design. The central composite design model and desirability functions could predict the best chromatographic conditions that achieved optimum resolution of the three analytes. The developing system composition was toluene: ammonia: methanol: acetonitrile in the ratio of (6.6:0.12:1.5:2 v/v/v/v) and the developed bands were scanned at 240 nm. The quadratic regression equations were selected for their higher correlation coefficients (r(2) > 0.999) and the detection and quantitation limits were in the ranges of (0.083 - 0.097 mu g/spot) and (0.253 - 0.294 mu g/spot), respectively. All validation parameters were investigated as per ICH guidelines and statistically compared using student t and F-tests. The results did not show any significant difference between the proposed and reported methods. Due to the short duration of the run, 10 min, the method can be used in the quality control routine analysis. modeling techniques were applied to investigate the inhibitory effect of 4-methyl acetophenone on the human carbonyl reductase enzyme (type I) and the potential health hazard which makes essential to control its level in the dosage form. The chromatographic conditions were optimized by applying quality by design principles during the development phase using experimental design. The central composite design model and desirability functions could predict the best chromatographic conditions that achieved optimum resolution of the three analytes. The developing system composition was toluene: ammonia: methanol: acetonitrile in the ratio of (6.6:0.12:1.5:2 v/v/v/v) and the developed bands were scanned at 240 nm. The quadratic regression equations were selected for their higher correlation coefficients (r2 > 0.999) and the detection and quantitation limits were in the ranges of (0.083 – 0.097 μg/spot) and (0.253 – 0.294 μg/spot), respectively. All validation parameters were investigated as per ICH guidelines and statistically compared using student t and F-tests. The results did not show any significant difference between the proposed and reported methods. Due to the short duration of the run, 10 min, the method can be used in the quality control routine analysis.