The Gateway theory (GT) proposes that tobacco or alcohol use lead to cannabis use, which can itself be followed by other illicit drugs (OID) onset. Aim of this study was to evaluate if the order of initiation sequence could influence further substance use. Data from a 2010 population-based survey were used (22,774 subjects aged 15-64). Using reported ages at initiations, 7 sequences were identified: initiation of tobacco only (T), cannabis or OID only, tobacco followed by cannabis (T-C), cannabis followed by tobacco (C-T), alternative 2-substance sequences, gateway sequence (T-C-OID) and 3-substance alternative sequences. Logistics regressions were performed to study the impact of sequence on further use (tobacco, alcohol, cannabis and OID), and substance use disorders (SUD) (tobacco, alcohol and cannabis). The most observed sequences were T (45.5%), T-C (20.5%), C-T (5.1%) and T-C-OID (3.5%). Further use and SUD likelihoods, whatever the substance considered, increased with the number of substances previously initiated. However, for a same number of substances initiated, current use and SUD likelihoods did not significantly vary according to sequence. Polysubstance initiation appears as a better predictor of further use and SUD than the initiation sequence, questioning the GT and being more in line with a common liability to substance use.