This study examines the determinants of informal payment practices (or “envelope wages”) in the European context, using a mixed-methodological approach based on machine learning and classical econometrics. Using the microdata of Eurobarometer 92.1 and the regional indicators of Eurostat which measure the same socio-economic indicators, this paper outlines profiles of people receiving some of their remuneration “under the table”. Following multiple imputation of missing values, the classes had been balanced using the ROSE function and then four supervised classification models were applied and compared: Logistic Regression, Random Forest, XGBoost, and Support Vector Machines. What is ultimately found, is that envelope wage recipients are not necessarily marginalised workers, but often formally employed people existing in a social environment where informalities of this type are considered normal. Important predictors also include having internet access, being financially vulnerable, social tolerance of undeclared workforce practices, and formal employment in specific sectors (i.e. agriculture and tourism). The highest performing model was XGBoost at over 97% accuracy with high recall, indicating this model will be appropriate for risk profiling policy interventions.