The use of online remote control for 24/7 behavioural monitoring can play a key role in estimating the environmental status of aquatic ecosystems. Recording the valve activity of bivalve molluscs is a relevant approach in this context. However, a clear understanding of the underlying disturbances associated with behaviour is a key step. In this work, we studied freshwater Asian clams after exposure to crude oil (measured concentration, 167 ± 28 μg·L−1) for three days in a semi-natural environment using outdoor artificial streams. Three complementary approaches to assess and explore disturbances were used: behaviour by high frequency non-invasive (HFNI) valvometry, tissue contamination with polycyclic aromatic hydrocarbons (PAH), and proteomic analysis. Two tissues were targeted: the pool adductor muscles – retractor pedal muscle – cerebral and visceral ganglia, which is the effector of any valve movement and the gills, which are on the frontline during contamination. The behavioural response was marked by an increase in valve closure-duration, a decrease in valve opening-amplitude and an increase in valve agitation index during opening periods. There was no significant PAH accumulation in the muscle plus nervous ganglia pool, contrary to the situation in the gills, although the latter remained in the low range of data available in literature. Major proteomic changes included (i) a slowdown in metabolic and/or cellular processes in muscles plus ganglia pool associated with minor toxicological effect and (ii) an increase of metabolic and/or cellular processes in gills associated with a greater toxicological effect. The nature of the proteomic changes is discussed in terms of unequal PAH distribution and allows to propose a set of explanatory mechanisms to associate behaviour to underlying physiological changes following oil exposure. First, the first tissues facing contaminated water are the inhalant siphon, the mantle edge and the gills. The routine nervous activity in the visceral ganglia should be modified by nervous information originating from these tissues. Second, the nervous activity in the visceral ganglia could be modified by its own specific contamination. Third, a decrease in nervous activity of the cerebral ganglia close to the mouth, including some kind of narcosis, could contribute to a decrease in visceral ganglia activity via a decrease or blockage of the downward neuromodulation by the cerebro-visceral connective. This whole set of events can explain the decrease of metabolic activity in the adductor muscles, contribute to initiate the catch mechanism and then deeply modify the valve behaviour.