Recent research has shown an association between monthly law enforcement drug seizure events and accidental drug overdose deaths using cross-sectional data in a single state, whereby increased seizures correlated with more deaths. In this study, we conduct statistical analysis of street-level data on law enforcement drug seizures, along with street-level data on fatal and non-fatal overdose events, to determine possible micro-level causal associations between opioid-related drug seizures and overdoses. For this purpose, we introduce a novel, modified two-process Knox test that controls for self-excitation to measure clustering of overdoses nearby in space and time following law enforcement seizures. We observe a small, but statistically significant ( p < 0.001), effect of 17.7 excess non-fatal overdoses per 1000 law enforcement seizures within three weeks and 250 m of a seizure. We discuss the potential causal mechanism for this association along with policy implications.