A simplified molecular input-line entry system (SMILES) is a well-known approach to representing the molecular structure of different substances. Quasi-SMILES is an extension of traditional SMILES by adding codes reflecting experimental conditions. The Monte Carlo technique of constructing and using quasi-SMILES for modeling nanomaterial toxicity is discussed. Examples of using quasi-SMILES for modeling nanooxide toxicity are presented. The influence of the so-called index of ideality of correlation (IIC) and correlation intensity index (CII) on stochastic Monte Carlo optimization processes is considered. The ability of the mentioned new statistical criteria of predictive potential to improve nanooxide toxicity models is studied. The advantages and disadvantages of the models calculated with quasi-SMILES are discussed. The possible perspectives of the approach in the aspect of the in silico simulation are listed.