The classifications of the hydrogeological functioning of karst systems used today are mainly based on indicators derived from spring hydrograph analysis. From these hydrodynamic approaches, it is sometimes difficult to make a link between the hydrogeological response and the properties or structures of the aquifer. To overcome this issue a promising way is to use the additional signal of electrical conductivity (EC), which provides information on mass transfer and on the complexity of the processes involved (transfer velocity, contributions from unsaturated and saturated zones, etc.). The objective of this work is to propose new indicators in order to improve the classification of karst aquifers based on a combined analysis of rainfall-discharge-electrical conductivity (P-Q-EC) time series. For this purpose, continuous hourly datasets of P, Q and EC over several years were collected on 21 French karst springs (SNO karst network and BRGM). A typology of these springs was carried out according to their hydro-climatic context, their structure (percentage of allochthonous recharge, importance of the saturated zone, etc.). In a first step, a statistical and signal processing analysis (frequency, correlation, and wavelet analysis) was used in univariate and bivariate approaches in order to produce different indicators giving insights on the hydrodynamic and hydrochemical responses of the aquifers. In a second step, a statistical approach was used to identify the indicators that differentiate karst aquifers according to their type. It was found that Q signal helps differentiate between karsts with a small saturated zone (Jurasian type) and karsts with a deep one (vauclusian type). On the other hand, the EC signal allows differentiating karsts characterized by a recharge from karst outcrops only (unary karst) from karsts influenced by an allochthonous recharge (binary karst). In a last step, a classification based on indicators derived from Q (memory effect) and CE (coefficient of variation) is proposed to differentiate karst aquifers according to their functioning and recharge type. Finally, this work shows the added value of the information contained in the EC signal combined with Q to better differentiate the hydrogeological response of karsts in relation to their properties and structures. This gives insights for classifications of karst aquifers accounting for vulnerability in addition to the groundwater availability estimated from Q signal.